Just imagine what you would do if your bread machine arrived without instructions. That of course, pales in comparison to the immense complexity of our brains (unimaginably more bake cycles).
Each of us happen to possess in our skulls, the most sophisticated computers ever conceived of and no one thought to provide instructions. No wonder changing how we do the simplest task, often meets with failure. If you climbed behind the wheel of a car for the very first time and had no instructions to guide you, how far do you think you'd get before driving into a ditch or up a telephone pole.
So, how do NLPer's create the knowledge necessary to learn how to operate our own minds?

Neuro-Linguistic Programming (NLP) studies the structure of how humans think and experience the world. Obviously, the structure of something so subjective does not lend itself to precise, statistical formulae but instead leads to models of how these things work. From these models, techniques for quickly and effectively changing thoughts, behaviors and beliefs that limit you have been developed. Warning! The following paragraph contains big, ugly, hard-to-understand words developed by a linguist. This is the only downside to NLP I know of and we're staying up nights trying to fix it. Parental guidance is suggested.

Many of the models in NLP were created by studying people who did things exquisitely well. Models such as meta-model, metaprogram, sensory acuity, Milton-model, representational systems and submodalities among others, provide a diverse set of tools for creating change in yourself and others.

Someone who wanted to create a model for learning to drive a car really well, might approach a expert in the field something like this - Instead of asking an expert driver, " How do you drive?" ("Very well, thank you."), they would be concentrating not on the content of what they did but on the underlying structure such as how they represent driving in their mind, the beliefs and attitudes they had about driving, the strategies they used in making decisions, how often they change their oil, (skip that last one) among other factors.

Let's use something called submodalities as an example of how a model works. By understanding how we perceive the world through our five senses, we can then understand how some people can respond very resourcefully in a situation and others do not. Once you learn how those who remain resourceful set up their representations, then it's a simple matter to teach others to do the same thing. The Example: Imagine seeing an enormous spider dangling directly in front of your face. Now clear your mind (sorry, didn't want to leave that image hanging around). A common way for people to have a phobic reaction to spiders or anything related to them, is to picture a spider completely oversized and far too close in their minds.

Spiders are tiny, well-mannered creatures that are far more frightened of you than you should be of them but try telling that to someone with that particular phobia.

So, why don't these phobic people notice the images they're creating? The popular belief is that we don't pay much attention to what's going on in our unconscious. If you considered the enormous amount of information your brain has to process each day, it's probably best that we don't spend much time dwelling on it (otherwise, we would probably sit around babbling and drooling and eventually starve to death).

Well, what do we do about our friend with the phobia, Extra-strength cans of Raid for a house warming gift?

NLPers ask the question, "If another person can have fun playing with their pet spider, what can we learn about them that we could teach the phobic person so they can play with spiders, too?" (Or something like that). The spider-lover would most likely have an image representing spiders that was proportionally correct and at a reasonable distance and possibly other factors not worth getting into right now. Knowing the difference, the NLPer can use one of many techniques to help the phobic person relearn their reaction to spiders so that it is similar in nature to the spider-lover's (hopefully less of the lover part).

The above example may sound complicated but phobia treatments often take less than half an hour. An powerful change with a minimal investment of time and effort.

NLP is based on many useful presuppositions that support the attitude that change is imminent. One of the most important is, NLP is about what works, not what should work. In other words, if what you're doing isn't working, try something else, anything else, regardless of whether what you had been doing should have worked. Flexibility is the key element in a given system, the one who is most likely to do well responds to changing (or unchanging) circumstances. That's one reason NLP has made so much progress in an area where such is not the norm. Innovators try out things with little regard as to its "truth" or "reality", NLP is much more interested in results and giving people what they want from life (sappy yes, but "true" nonetheless).