Does anyone know the history or rationale of the 12 inch limit for length as stated in article 350-18? TIA

I was taught that the 12 inches was about the shortest length of flex, entering a box, that could be bent in a 90. Consider the amount of flex that goes straight in to a flex connector, plus the flex that is needed to bend a 90, then strapping it..roughly 12 inches.
Course, this rationale doesn't work for larger diameter flex. The larger sizes need about 12 inches just to straighten an offset out before strapping it.
Just to add to the thread: Is the 12 inches measured ALONG the flex (or MC, AC, or NM), or just the distance from the box connection to the support point?

Thank you both for your prompt reply. As chairman of the electrical safety committee at a national lab, I have been asked by some of my colleagues to evaluate a proposed wiring method to be used in new construction of residential units using compact fluorescent downlighting fixtures.

My colleagues have been working closely with the construction and lighting manufacturing industries on a national level as well as UL. The design incorporates some of the features of article 604, Manufactured Wiring Systems. A 'plug and play' lighting system if you will. However one major drawback that I see is the 12 inch limit from last place of attachment, a ceiling joist in their design, for the FMC. Optimally, their design needs about 18 to 24 inches of FMC length from the joist attachment to the fixture. I have ruled out the 3 exceptions under 350-18. I have convinced them that their design should not rely on an exception but should stand alone and pass judgement by an article of the code, not an exception. This would greatly reduce the various interpretations that may arrise by the various AHJ's.

At one time, I'm sure there was justification for the 12 inch limit and that it was not an arbitrary length pulled out of the air. I would be very much interested to know what the justification for the 12 inch length was. Regards, Art

Boy, I'd love to have a book about how all the mystery dimensions were arrived at!
My oldest 'code' book (by "National Board of Fire Underwriters" --"recommended" by the NFPA), is 1959 vintage. And the language was extant back then, borrowed from the section for AC cable.
The engineering issue is one of passing UL testing for the integrity of the connection, and not over-bending and damaging the flex. As a "manufactured system", perhaps there is no need to worry about the NEC, as long as the design 'works' for the application..with due regard for 'idiot proofing'..not an easy task!
If no one here supplies you the info you need, but you find it elsewhere, I would sure appreciate hearing from you about where you find such historical info.