I think it's likely that the MBTI preferences represent what you might call social niche strategies, and that the current percentage of N's (for example) represents some kind of optimum level of N's within the population — or at least, what used to be the optimum level, during whatever evolutionary years are responsible for the current mix.
Given enough time, evolution is a very powerful process. A genetically-caused trait that confers just a 1% reproductive advantage will, over the course of many generations and in the absence of complicating factors, gradually become a trait present in virtually every member of the population. So for example, if extraverts had been more reproductively successful than introverts for most of our evolutionary history
no matter how large a percentage of the population was made up of extraverts, we would presumably all be extraverts today.
I suspect that the fact that there are substantial numbers of people on both sides of each temperament dimension suggests that, any time the percentage of the population exploiting one niche over the other becomes larger than optimal, it becomes advantageous to be a person who naturally exploits the other niche — with the result that, over time, a more or less optimal balance among the types is maintained.
Jung speculated that introversion and extraversion corresponded to competing reproductive strategies, each successful in its own way:
Jung said:
There are in nature two fundamentally different modes of adaptation which ensure the continued existence of the living organism. The one consists of a high rate of fertility, with low powers of defense and short duration of life for the single individual; the other consists in equipping the individual with numerous means of self-preservation plus a low fertility rate. This biological difference, it seems to me, is not merely analogous to, but the actual foundation of, our two psychological modes of adaptation. I must content myself with this broad hint. It is sufficient to note that the peculiar nature of the extravert constantly urges him to expend and propagate himself in every way, while the tendency of the introvert is to defend himself against all demands from outside, to conserve his energy by withdrawing it from objects, thereby consolidating his own position. Blake's intuition did not err when he described the two classes of men as "prolific" and "devouring." Just as, biologically, the two modes of adaptation work equally well and are successful in their own way, so too with the typical attitudes. The one achieves its end by a multiplicity of relationships, the other by a monopoly.
A June 2011
OpEd piece by Susan Cain in the
New York Times described an experiment involving impulsive and cautious fish ("rovers" and "sitters"), and here's a bit of it:
Susan Cain said:
We even find "introverts" in the animal kingdom, where 15 percent to 20 percent of many species are watchful, slow-to-warm-up types who stick to the sidelines (sometimes called "sitters") while the other 80 percent are "rovers" who sally forth without paying much attention to their surroundings. Sitters and rovers favor different survival strategies, which could be summed up as the sitter’s "Look before you leap" versus the rover’s inclination to "Just do it!" ...
In an illustrative experiment, David Sloan Wilson, a Binghamton evolutionary biologist, dropped metal traps into a pond of pumpkinseed sunfish. The "rover" fish couldn’t help but investigate — and were immediately caught. But the "sitter" fish stayed back, making it impossible for Professor Wilson to capture them. Had Professor Wilson’s traps posed a real threat, only the sitters would have survived. ...
Next, Professor Wilson used fishing nets to catch both types of fish; when he carried them back to his lab, he noted that the rovers quickly acclimated to their new environment and started eating a full five days earlier than their sitter brethren. In this situation, the rovers were the likely survivors. "There is no single best ... [animal] personality," Professor Wilson concludes, ... "but rather a diversity of personalities maintained by natural selection."
Cain is the author of the best-selling
Quiet: The Power of Introverts, and I'd say she errs in framing the rover/sitter duality exclusively in E/I terms. The most well-established Big Five test (McCrae & Costa's NEO-PI-R) breaks Conscientiousness — i.e., J/P — down into six "facets," and one of those facets is called
Deliberation. As McCrae and Costa explain: "Deliberation is the tendency to think carefully before acting. High scorers on this facet are cautious and deliberate. Low scorers are hasty and often speak or act without considering the consequences. At best, low scorers are spontaneous and able to make snap decisions when necessary." And decades of both MBTI and Big Five data have pretty clearly established that E/I and J/P are essentially independent dimensions of personality.
I'd be more inclined to suspect that the J/P dimension and the neuroticism dimension (which includes anxiety-proneness) are also meaningful contributors to the rover/sitter duality — in people, at least — but I'd also say that viewing E/I as a contributor is consistent with most MBTI sources and, in any case, I think it's probably fair to say that the most impulsive, bold, plunge-right-in types are the Calm EPs and the most cautious, look-before-you-leap, think-before-you-speak, worry-prone types are the Limbic IJs (like me).
In any case, and as a final note, there's lots of room for reasonable people to argue about both the exact
nature of the major clusters of human personality variation and the causes of those clusters. But arguing that there's
no such thing as substantially hardwired variations in personality, and/or that such variations would, ZOMG, be
inconsistent with evolution, is just silly at this point, given the existing data. Decades of twin studies strongly suggest that genes account for around half (or more) of the kinds of relatively stable temperament dimensions measured by the MBTI and Big Five. Identical twins
raised in separate households are substantially more likely to have similar personalities than less genetically similar pairs.