I'm more inclined to see people as NF, NT, SF, ST myself. I've never understood why it was SJ/SP. Why would it be one way for the Ns and another way for the Ss? I relate to the STs in an entirely different manner than the SFs, and I think the F/T makes that initial distinction first and foremost, well before the J/P. Just what's bugged me, I guess. Then again, I'm not a scientist.
Well, Keirsey is INTP (I think) and a theorist, so much of his work is constructed from the theory side of things.
I think he is grouping off motivations, primarily, here.
Concrete J's (SJ) = Si as first or second function
Concrete P's (SP) = Se as first or second function
Apparently to Keirsey, he saw this as creating more of a separation than the T/F difference. He was looking at "motivations," and concrete J's are more interested in creating the inner map of the world and changing the OUTER world to conform, while concrete P's are more interested in responding to the outer world in the moment -- they change THEMSELVES [i.e., their actions] to accommodate and overcome what is happening in the outer world.
One imposes on the outer world, one responds to the outer world.
<Sorry if this is not going anywhere, I have never thought in this direction before... this is very raw!>
Now if we did this with intuitives, what happens?
Abstract J's (NJ) = Ni as first or second function
Abstract P's (NP) = Ne as first or second function
Is there a useful delineation here? Probably there still is... except that N's generally are more "free" like the SPs are, they have flex because they're looking at possibilities. So perhaps they are harder to distinguish until you begin to look at HOW they interact with the world and what they try to change/enact? (This is why T/F perhaps seems more influential with N's.)
Sorry, I'm losing my train of thought and need to get back to work, but please feel free to take the ball and run with it.
Yes, I've read the Keirsey book and saw how that classification system worked. (While it was fun to read, most other parts of it, like the "directive vs. informative" or different types of intelligence, seemed way to symmetrical to be useful for much.)
There is that problem -- it balances out nicely on paper... but does it accurately reflect reality in the most coherent way? I don't know.