So, yeah. We've got dominant and auxiliary functions, which types are known for. Some folks like to drag in tertiary and inferior functions as well (often as an 'argument by my type knowing better than yours because you have inferior Fe, haw haw haw'; but I digress).
What about those poor, poor shadow functions? Our perceptions of how those are supposed to play out are all over the place.
Some regard, say, INTJs as not 'having' or 'using' Ti because they use Te instead--that they synthesize Ti with a combination of Te and Se (or something). Some claim that Ti for them would be in some 6th position that means something-or-other; I forget exactly.
Others, like me, believe differently and invoke time-proven principles such as Occam's Razor and 'if it walks like a duck and quacks like a duck then we may as well call it a flippin' duck' and that, screw it, INTJs can and do 'use' Ti.
So, when we talk orientations to different functions or treat functions as axes (e.g. SeNi axis versus SiNe axis), we may as well get a bunch of ideas out there and see which ones stick.
Personally, I regard most of the function position talk as unnecessary and counterproductive. I'm of the belief that the fluff consists of any discussion past the dominant function, a smidgeon of auxiliary, and a mental note that we innately repress the opposite of the dominant. Defining function positions more rigidly lead us to describe complex mental phenomena with a structure composed largely of toothpicks and chewing gum--it might hold up as adequate for a 3rd grade science fair, but not for a proposal to a panel of keen psychologists.
If we're to assign roles for all of the functions, I like the way Socionics does it. It actually acknowledges that all types 'access' all functions, even those that are counter to the types' egos.
What say you?