The notion that if you're an "Ne type," you're also a "Si type" — and ditto for the Ni/Se, Fi/Te and Fe/Ti pairs (the so-called "function axes," or "tandems") — is a by-product of the Harold Grant function stack, which is the forum-famous model that says that INFPs are Fi-Ne-Si-Te, and INFJs are Ni-Fe-Ti-Se, and ZOMG, INFPs and INFJs have
no functions in common!
And just so you know, that model is inconsistent with Jung, inconsistent with Myers, and has never been endorsed by the official MBTI folks. More importantly, and unlike the respectable districts of the MBTI, the Grant function stack has no substantial body of evidence behind it — and should probably be considered all but
disproven at this point.
Here on Planet Reality, the fact is that the four dichotomies, not the functions, are the real, underlying (and substantially genetic) components of your MBTI type — and despite some Jungian lip service, Myers understood that, based on her years of data-gathering and psychometric analysis. And she also understood that
dichotomy combinations were associated with many noteworthy aspects of personality, but that there was nothing particularly special about the combinations that are purportedly associated with the "cognitive functions." In fact, Myers thought of NF/NT/SF/ST as the most significant dichotomy combinations — and it's worth noting that that's a carve-up of the types where each group is a type foursome with (assuming you believe in the functions at all)
four different dominant functions.
And you will search in vain for any passage in Myers where she says that, if you start with a type foursome that shares two preferences (e.g., the SJs), and you flip both preferences, you'll end up with a foursome (in this example, the NPs) that has
more in common with the original group — when it comes to some or all of the stuff affected by those preferences — than if you'd only flipped one preference. And the reason you won't find any such passage is that Myers didn't subscribe to that notion at all. Myers understood that if there's an aspect of personality where the SJs are the types with the
most of it, you should expect the NPs to be the types with the
least of it.
And Myers was right. The HaroldGrantian double-flip — the goofball geometry underlying the so-called "function axes" — has no basis in reality, and that's why it's found
no respectable validation in over 50 years of MBTI data pools, correlating the types with everything under the sun. The notion that an INFP has "tertiary Si," and will therefore tend (
probabilistically speaking) to have "Si" aspects of personality in common with a typical ISTJ that ISTPs tend
not to exhibit, is a typological assertion that — like all assertions that
crosscut the dichotomies in that counterintuitive way — has no more
validity than the notion that two people born at around the same time will tend to have aspects of personality in common because they're both Capricorns.
In case you're in the mood for a hefty helping of input on the relationship between the dichotomies and the functions, the place of the functions (or lack thereof) in the MBTI's history, and the tremendous gap between the dichotomies and the functions in terms of scientific respectability — not to mention the unbearable bogosity of the Grant function stack — you can find a lot of potentially eye-opening discussion in
this post and the posts it links to.
The final link at the end of that linked post is no longer functional (since the owner has taken INTJforum private), but you can find a long replacement excerpt from the INTJforum post — describing the dichotomy-centric history of the MBTI — in the spoiler in
this post.