I first read about Parts in Tranceformations. It made so much sense to me that I went and bought Reframing and played around with that for a while.
Then various people would claim that Parts was an outdated model yet wouldn't explain why (merely that Richard Bandler no longer used it).
I couldn't find anything that Richard had written about it but I did find
this article
by Carmine Baffa.
Just because your mind can ACT as if you have a part, doesn't mean that's
a good way to organise your whole brain.
I don't like giving my clients any more limitations than they already have. Therefore I generally don't communicate with them through a model which presupposes any particular limitation, such as their mind/body not being one coherent system.
See, if you have two parts with separate intentions, they're almost guaranteed to conflict. You want to struggle in life? Install lots of parts!
But, in much the same way as reimprinting, the Parts model can be very effective at dealing with the presented problem. It is very common for people to stick with what they know that works. And since an ineffectual intervention performed confidently will often work better than stumbling through a more appropriate intervention I can understand people's reluctance in abandoning their over-reliance on the model.
As such, using the Parts model may be the best choice you can make in the short term, but realise that in the long term you can offer better models to your clients.
The main excuses generated by the left hemisphere of these reluctants are:
1. Because the model is a metaphor, it will not change the way the client
structures their mind.
Yet Richard Bandler and Milton Erickson have often installed new patterns
directly into the unconscious mind using metaphor exclusively! This excuse
also seems ignorant of the NLP presupposition "The meaning of your communication
is the response that you get".
2. The model accurately represents the structural nature of the brain.
Simply isn't true. No longer do we talk about the limbic system, we talk about the limbic region as it has many projections into the cerebral cortex. The model does not encourage intracommunication necessary for optimal mental health. Plus you can simply talk about the functioning of those aspects of the brain. Pull out a neurology book with pretty pictures if that helps. There's less chance of the client generalising limitations than with the Parts model.
Perhaps it is John McWhirter who has made the strongest criticism of the Parts model, calling it a conceptual error of (neuro)logical levels:
"The identification of separate "parts" of a person is a confusion of value (like to, want to, need to) with identity (is, my, our, name, thing). This is a common error, which is identified in other areas of NLP but was never applied here."
However, the potential appeal of the model can get you rapport and give you leverage very quickly where that is necessary.