Socionics is a theory that is based around the original idea, or model A. But it’s not quite that simple, there are other models as well, adding different aspects to the theory of socionics.
Model A is the basic socionics, coming from Aushra Augusta, the founder of socionics. Intertype relationships, the function stack, what the functions mean, etc. This is the foundation, basic socionics, that everything that socionics is revolves around. Accepted by everyone that is working with socionics.
Model B is a model where they included signs for each function as well. Instead of using each function in isolation, there are two different aspects of it, one external and one internal. In for example an LIE, the dominant function isn’t Te, but Te- Ti+, both holding certain aspects of each function.
(for more information about model B, click here and here)
This model goes back to using only four functions, like Jung did. There’s also an introduced concept of excitation and inhibition, and how those are applied to the function stack. Excitation and inhibition are applied differently for introverts and extraverts. For more information about model T, and the pattern of excitation and inhibition, click here.
The most recent of the models. It’s focused on energy rather than information, in contrast to conventional socionics, and because of that, uses a different stack than model (however, it covers two different things, and is therefore not in contrast with it). There are four different “levels” of energy. The most is given to our leading and creative function. Our demonstrative and suggestive function gets a a little below that. After that comes our vulnerable and role function, and the least energy is focused on the mobilizing and ignoring function. The functions are also divided into “externalies” (the functions of the preferred attitude (E or I)), which are focused externally, on society, and the “internalities”, focused internally, on ourselves.
Grigory Reinin introduced an additional 11 dichotomies, besides extra- and introversion, sensing and intuition, thinking and feeling, and rationality/irrationality. Three of those are central in understanding the difference between quadras, and are already described shortly in my post on quadras.
Visual identification is not really a different model, but a typing method. I include it here however, because its use is largely disputed. The idea is that there are visual clues to a person’s type. These clues can come from body language, our choice of language, or in the structure of our faces. Most often though, even in the case of the face, it’s not the genetic factor that is the important part, but rather different expressions. It’s mostly seen either as the most reliable method of typing people, or completely disregarded.