Personality type may not be a direct phenotype of any certain allele inherited from ones parents, but may be a byproduct of combination's of multiple alleles. A certain brain structure may develop because of our inherited genetic makeup, and the personality type may only be a result of these phenotypic physical affects on our brain structure.
I suppose one would have to find out the combination's of alleles at the different chromosomal loci - which would require defining how large of a loci to look at - to know how many alleles combined would create a brain structure that makes one "INTP" or "ESFP" or whatever else. That's beyond the scope of this reply, though.
Let's just say that each trait - Ne, Ni, Se, Si, Fe, Fi, Te, Ti - each has a different allele in their respective loci in the genes that code for our brain structure. The functions themselves are byproducts of the way our brain becomes structured according to this genetic coding. The sixteen types of brain structure evolved either from many active germ-line lineages passed down from our ancestors, or some of them may be emergent mutations from more recent times (it's difficult to say if INTP's existed 30,000 years ago). Natural selection would have found these specific brain structures more economical and advantageous, thereby evolving a set of alleles that are found predominantly throughout society (that is, if MBTI is true) - but that would not mean new mutations could not arise (some MBTI types might be older then others).
So why would some types be more rare then others?
Perhaps some alleles are recessive (possibly iNtuitive or Introverted alleles) and others are dominant. This would make it more difficult for children to be born as iNtuitives or Introverts, as they would require both recessive genes from their parents to be an iNtuitive or Introverted. This would also mean that each of us carries a gene for a type that isn't being expressed (I wouldn't confuse this with shadow functions, though) that we may pass on to our children. This would explain why a lot of people on this forum seem to have been conceived by two "S" and/or "E" parents (they simply might just be 'carriers' of the correct alleles for INTPness).
Perhaps some types are not in decline, but are a more recent mutation. The Introverted lineage certainly doesn't seem to have a lot of fitness, speaking in evolutionary terms, and it would not be as advantageous to be an Introvert as it would to be an extrovert if one were looking to pass on those genes (whether they are recessive or not).
It could also be a matter of balance - possibly brought upon by traditional cultural norms. If in a society of humans the number of introverts begins to rise, then the extroverts, who in America at least are seen as more attractive (or at least more aggressive about finding mates), begin to reproduce more often, since they will be more heavily saught after. If the number of introverts begin to wane, it would be due to reproducing with extroverts, which would mean a lot more extroverts with recessive introvert genes that would then be passed down to the next generation, making more introverted children (since more double extrovert couples would have the dormant recessive introvert genes). This would be a constantly fluctuating system (which makes more sense then a static percentage).
Just some thoughts - feel free to start a
formal debate with me if you disagree!
.
