Wheel IM model
picture of the model:
wheel_IM_model.png
Overview & Basic Premises
The Wheel model of information metabolism (IM) is an attempt to fix many fundamental metaphysical problems in model A and its offshoots. Briefly, this model will define one unified, circular data flow & place all 8 functions at fixed positions within this flow, where they all actively participate. To create this model we start by defining its metaphysics. We then start to describe the information metabolism. Information metabolism is the process by which information transforms as it passes through IM elements - a subject takes in raw sensory data, processes it, stores it, recalls it, represents it, and restructures the environment with it. Data flow is a related term for the directed graph of data passed between IM elements. The starting point for our model of the data flow can be represented by a ring. We progressively divide the flow up into portions, often called IM elements. You can see these divisions physically represented by the dividing lines in the image. Each point where the dividing line makes a cut through the circle (the flow) we will refer to as a data structure. Points like these are representative of data in a particular structural configuration at a given time. The space between one point and the next will be called a phase. Functions are an example of a phase. A phase has a data structure as an input, and a different data structure as its output state - data transforms as it passes through the phase. The term IM element will be used to refer to either phases or data structures interchangeably. IM is all about state transitions, where state is an arbitrary configuration of information.
Dividing up the data flow is done in steps, and results in an IM element hierarchy. For example, drawing a line through the circle once produces 2 points / 2 phases. Cutting it a second time results in 4 points / 4 phases, cut it a third and fourth time and we have 8 points / 8 phases, etc.. This would have created 2 top-level IM elements to start with, followed by 4 mid-level IM elements (children of the 2 top-level elements), followed by 8 bottom-level IM elements (note that the phases at this level correspond with the 8 functions), etc..
Notice that IM elements have fixed positions within the flow relative to one another. This is one major way our model differs from model A. In our model an IM elements definition is derived from its physical position in relation to other elements in the hierarchy we've created, and in the flow. The flow is circular and consistent - information always flows in one direction (at least in the basic model this assumption holds - later we will talk about how very temporarily the flow can reverse, but that's an exception to the general rule). You could think of the undivided data flow as the transcendental structure on which the rest of the model is based. This underlying structure is something model A sorely lacks. Many fundamental problems in model A are rooted in its lack of sound metaphysics, but we will go into that.
There are more nuanced aspects of our model. A person can be mentally fixated on, or favor, a particular IM element in the flow. This is akin to having a particular view of the world, and is what we'd call a persons type. Type here is a more flexible concept than in typical socionics - a person can be identified by an IM element at any level in the hierarchy, though the further down the hierarchy you go the more specific the identifying traits become. There is also an idea of strong or weak IM elements related to the angles between IM elements in the flow.
The flow can be divided up into any arbitrary number of data structures / phases, and assuming that they have a well-defined hierarchical relationships this model accommodates an endless number of alternative theories, bringing them together in one model (something no other socionics model can do...).
This post will be updated with information pertaining to this model, explaining its basic premises, describing the phases and data structures on a more mechanical level, defining what a type actually is, covering aspects, the possibility of multiple types and multiple independent data flows, and other things
Rational idealism, why socionics must remain a rational idealist model
A pure idealist assumes everything that exists is in some way mental... a weaker idealist may just assume that the only way we can apprehend that which exists is by a mental model of it, we can't apprehend it directly. Idealism isn't concerned about tangible things, rather the way those things are perceived or represented. You could say that interpretation, for an idealist, is a feature of the mind, and not primarily determined by the thing in reality... or at least that we should give priority to the mental over the non-mental. Now, obviously that philosophy isn't entirely correct, it doesn't directly contend with material reality (or in our case the brains physiological structure). The rejoinder to this is transcendental idealism, which claims that, while the mental model is an ideal representation, the model suggests something that's true about the material reality. The mental model could be reconciled with the reality, and mapped to it. A transcendental idealist categorically assumes some sort of correspondence between their ideas and reality.
For a rational idealist the assumption of this correspondence is justified by an appeal to reason. If the idea makes sense it ought to map to reality in some form of fashion, even if this mapping requires some further elaboration or development of the idea.
For example... we have a system of math that justifies the statement 3+3=6. Now math was created with pure reason, not with observation... math is an ideal rational representation. Lets say 3 ducks walk by... and around the same time I look off into the distance and see 3 cows grazing in a pasture, and I can say that there are 6 animals. And I've taken this ideal model and used it to reason about the world.
Some materialist philosophers will argue things like... there is no physical object in existence that is a perfect square... so math is always a bit wrong. When I say we have "3 ducks" there are actually differences between the ducks and I can't just group them into the same category and add them together. And I can't actually add cows with ducks either, that doesn't make sense.
But again, all thought is idealistic, math is useful because it aids in reasoning and allows us to make predictions. The differences between the ducks, the fact there is no perfect square, we just ignore that while we engage in the reasoning process.
But furthermore, reality is logical, even if the complexity of the physical world makes it hard to comprehend all of the logic. And so regarding these little problems in quantifying a duck, the fact there is no perfect square or straight line that exists - we must take for granted when we engage in reasoning that the categories of reason and reality as it actually exists can be reconciled.
And the mental also has some inherent value.
Anyway, the same is true with socionics. Socionics is the same sort of idealized model... it was created with pure reason, it was not meant to be a representation of some physical brain structure. However, despite it always being a bit "wrong", or even very wrong, it is useful... it can be used to describe thought in any context... this makes it very useful like math, but when we use it we just assume that there is some possible reconciliation between our reasoning and reality. And this is why we can use it to describe almost anything we like - because what we're describing is the thought that pertains to the particular thing.
You can criticize rational idealism, there are many philosophical criticisms of it... I think the criticisms are a bit meaningless since human thought, and reason itself, is always going to be somewhat idealistic, even language itself is idealistic. This criticism would actually lead you away from doing socionics. The very concept of a function is idealistic, as is the notion of an IM. Every model in socionics is grounded in these ideas, and justified by them, even if they don't acknowledge it, even if they only refer to the Jungian functions and completely discard model A...
Model A was created using pure reason. At the core of it is a set of functions and an information metabolism. There is a metabolic flow from one function to the next. Some theories are justified by observations, sometimes combined with statistical methods (like the big 5), but socionics is not one of these. We aren't gathering observations and basing our models off them... at least not in any systematic way that would substantiate the model.
Many theoretical offshoots of socionics have abandoned the rational idealist basis of socionics, and attempted to turn it into a taxonomy system that takes things such an understanding of the brains physiology into account when designing the model. For example, some socionics models claim the IM only applies to the inner workings of the brain, and redefines the sensing functions to disclude any mention of the sense organs.
These taxonomy systems have not followed any rigorous rational or scientific method in their creation, and as such will always categorically remain pseudoscience. Socionics, from the original conception of model A, has always aspired to more than this. For example, if our goal was to create a simplified model of cognition that accounts for the brains physical workings we'd need to derive the model from the brains physical workings, not from some abstract notion of a cognitive function, or an information metabolism. There are neuroscientists who have come up with theories about how the brain works. How the brain works still remains a mystery, because we remain unable to properly define consciousness and how it operates on a physiological level. There also appears to be a non-local aspect to consciousness, and the brain is also subject to the laws of quantum physics and other kinds of physics we haven't completely developed... but that's another conversation. With socionics we set aside the physiological details and attempt to model thought directly using reason, that is a core assumption of modeling cognition as a functional metabolic process.
To establish arbitrary boundaries between the sense perceptions and internal cognitive processing you're forced to appeal to neurophysiology. How does neurophysiology relate with the socionics model of cognitive functions and information metabolism? We have no idea, we didn't develop socionics by looking at the brains anatomy. There is some overlap between the socionics model of cognition and how the brain physically works, some sort of mapping.... but we have no way of actually knowing what that mapping is. Assumptions about how this mapping occurs are arbitrary, just pure guesswork. And we don't really need to make some attempt at doing that mapping, it's just a core assumption of any rational idealist model that this mapping can occur, and this is justified by reason. Furthermore, by restricting the cognitive functions to describing just the internal cognitive processing you don't actually reconcile your ideal model with the brains physical structure, you've just created a model that's half physiologically based and half rational idealist. So now the model has just become incomplete and incoherent, you've made no progress in reconciling the two models, this schism between your theory and reality remains.
So our model deliberately makes no attempt to model brain physiology, it aims to model cognition directly using pure reason. The appeal to reason is the strongest philosophical justification for itself socionics has, without this it has no argument or evidence to suggest its models are correct. The reconciliation between our model and the brains physiology is taken to be implied by reason.
Our models begins by reasoning about metaphysics and from there develops into a coherent system.
metaphysics and the need for a transcendental form
The goal of designing a generic rational ideal model like socionics is to create a structure that is malleable such that it can accommodate any kind of content in its domain. Socionics is fundamentally interested in the form of thought. There are other similar models in different fields - grand theories of physics, language, etc..
To achieve this the model needs to be based on sound metaphysics, which inevitably means accurately describing or modeling the transcendental structure of reality. This is something model A makes no attempt to do.
Metaphysics can be established through a process of continuous abstraction. A metaphysical model is interested in modeling a state of utmost potential, but one which is devoid of content.
Arbitrary assumptions inherent in a model - its tangible content - make it inflexible and limit what can be modeled. Such assumptions are easily, and inevitably undermined through the reasoning / abstracting process, and through this process a metaphysics is established.
After all content and structure has been abstracted what remains is a transcendental form.
The transcendental form is hard to describe in tangible terms since it transcends and permeates all content and structure. There are no distinctions by which to easily define it, but we can say it acts as a well of infinite potential, it is characterized by its flexibility and lack of distinctions, and that it unites all structure together coherently.
There are some well known ways of modeling the transcendental form, and many of the symbols used for the model are common occult symbols. One such way is as a single orbital self-referential point. This is what the symbol of the ouroboros actually represents. Another way of modeling it is as a spiral. The spiral emphasizes the non-repeating nature of transcendence. But for our purposes we will just model transcendence as a simple circle - our information metabolism ring. This is a directed graph. Its data flow is circular (this is vaguely why the model was named "wheel", to refer to this underlying circular structure). Unlike in model A, our model is coherent - every type can be modeled with it.
All IM elements are derived from a transcendental form (a process we will discuss in detail), forming a hierarchy of IM elements. In the process elements have a set unique relationships established with one another within the hierarchy, and with the transcendental form itself. We will refer to this as an IM elements transcendental aspect. This transcendental aspect is entirely unique - it uniquely characterizes and identifies the element. An IM elements traits and attributes are all part of its transcendental aspect. Thus an IM elements traits do not change (it would completely undermine its definition if they did), and are defined holistically. There will be nothing arbitrary about the traits and attributes of IM elements in our model.
Another way of phrasing this is to say the data structures / phases are ultimately defined by both their relationships with other data structures / phases at various levels, and in what they produce on the whole.
The transcendental form can be differentiated infinitely, and we can derive an unlimited number of IM elements from it, with increasing levels of specificity and complexity.
We can't ignore the transcendental form, it is what ultimately defines elements and relates them all with one another coherently. Model A ignores it, but Model A is incoherent (we will discuss how in a bit). Without a transcendental form a theory has no broad context by which to uniquely identify elements and define them. Such a theory is unable to distinguish superficial traits from core ones, or remain consistent in definitions.
Even the very notion of a circular data flow relies on an underlying transcendental form. Without some sort of transcendental aspect nothing will persist between data transformations, because the data structures have nothing in common. Thus a transformation between the data structures can't occur, and certainly can't be reversed - therefor there can be no circular data flow.
The transcendental form is tautological in that every IM element that can be derived from it will always arbitrarily and coherently relate with every other derived IM element in the hierarchy.
Note that in our process whenever we derive an element we create an opposing element (every cut creates 2 points). This is just a limitation of the binary process we chose to use, but the model need not be strictly bound to this process. So long as the resulting phases are proportional the model is sound. For example, a single self-referential orbital point is a well known and fundamental metaphysical model (i.e. the ouroborus). But for our purposes a binary works.
The self-referential orbital demonstrates there's a temporal element to the model. This is the basis for the concept of data flow. The concept of data flow actually implies that some particular piece of data - which we could consider a self-referential orbital point - is located somewhere along the orbital continuum of our model, in one of its phases. In this sense, each IM element in our model is actually a description of the same piece of data at that particular point in the data flow.
In formal logic a tautology (such as the logical circularity in our model) logically proves that a claim is true. It's clear that no other model is sufficient for our purposes.
Why Model A is incoherent
If a general model such as socionics is not properly grounded in a sound metaphysics it will end up forming superficial, niche, or faulty assumptions. There will also seem to be no way of critiquing its assumptions, they will seem very arbitrary. In turn this can undermine everything in the model from definitions, relationships, concepts, the way data flow works, types, and so fourth. It may even result in a model that is incoherent, and perhaps does not even recognize the need for coherence.. Unfortunately in Model A this very much happened. Model A is not just a superficial or niche model, it is completely incoherent.
One of the most critical flaws in model A is the importance it places on whether a function is introverted or extraverted, combined with whether it's a perceiving or judging function. This was the original theoretical justification for pairing functions together into function blocks. The assumption is that the output state of a perceiving function must be the input state of a judging function, and vice versa; and that likewise the output state of an introverted function must be the input state of an extraverted function, and vice versa.
In reality a function is always defined in its entirety by its input / output states. These states are characterized by all of a functions information aspects, not merely a particular aspect. As such a functions input / output states are unique to it.
For example, most theoreticians today agree that the external / internal aspect is more significant than the introverted / extraverted aspect - Ne and Ni, for example, are both internal functions, and they are processes that occur within the subject; whereas Se, in contrast, is an external function and occurs outside the subject - with Se the subject is taking in the raw external information. The input and output states of Se are simply not interchangeable with Ne, they are worlds apart on account of the external vs internal information aspect.
Furthermore, a function always accepts the same input and produces the same output, because this is what defines a function. To change a functions input / output is to change the function itself. Functions in the IM are not like legos. In pairing functions together in this blocklike manner Model A presumes to actually change the input and output states of functions for different types. Model A has thereby inadvertently created 8 unique functions for each and every type, shattering any semblance of coherence in the model. i.e. the Ne that couples with Ti is not the same Ne that couples with Fi. This leads us to define no less than 128 unique functions. But Model A does not actually define these functions, or even acknowledge its model necessitates this. Nor could it ever justify doing this in the first place. The types in model A have no functions actually in common with one another, and the information metabolism is completely incoherent between them.
Ultimately there is just no justification for model A's concept of function pairings, it violates the very idea of what a function is.
These function pairings are a cornerstone of Model A, and without them the entire model falls apart.
Functions do not change their input output states. By logical necessity when any organism uses Se (takes in raw external information) it must sift through that information and interpret it, determining what information is useful and what information is useless. The subject always performs that interpreting / pruning function, in some form or another, before it commits the information to long memory. This metabolic process exists by logical necessity. You can see it in the brains structure. It even is present in AI models. There is no version of Se where the subject does not need to perform some function of interpreting and pruning the information to prepare it for deeper assimilation. Se always, by definition, is followed by Ti in the metabolic process. Both Se and Ti are external static functions. The output state of Se is the input state of Ti - what distinguishes them is one aspect, the object / field aspect. They are part of the same higher level phase. This can never change, because it is part of their very definition. We talk about this more later.
Another issue in model A is it has no consistency in the directionality of functions. For example, model A's function blocks are bidirectional - SLE uses Se as their primary function, and Se's output state feeds into Ti as its input state; but for LSI the reverse is true, Ti's output state is Se's input state. In reality when a subject perceives raw information (Se) there's an abundance of it, and by logically necessity they must interpret and prune the information. They must identify what parts of the information are logically pertinent to what they're interested in. This is an operation handled by Ti. Like Se, Ti is an external static function. This process is not reversible for some types - interpretation / pruning cannot immediately precede raw perception, this is not a cognitive style, it undermines the very notion of the operation and is logically incoherent. To prune information... raw information is required. Again, the input / output states of functions always remain consistent - functions don't work like legos where we can swap functions around and piece them together, this alters the input / output and changes the function itself.
These are a few of the fatal flaws in Model A. There have been repeated attempts to rectify these problems, but no model has yet uprooted model A and recreated socionics from scratch. The problems in model A are so extensive they are beyond repair. Augusta had no concept of metaphysics, there is no principle of coherence at the core of socionics, she was just gluing ideas together and going with them.
Deriving the IM elements
The subject & object
The subject & object are the two most basic data structures in our model. Information presupposes a subject & object - the perceiver and that which is being perceived. The Wheel IM models how information transforms as it passes between the subject & object in a circular manner. Raw information in the environment is perceived by the subject... the information is sifted through, interpreted and pruned. The information is then evaluated in relation to the existing value structure. Certain information is retained, and committed to memory. Memories may arise spontaneously, or something might trigger them. The subject must remember that which is relevant in order to operate effectively in the world and express themselves in it. Recalled information must be transformed into a presentable or useful form. After that the subject can use the information to restructure their environment. The way the environment is structured to the subject guides their raw perceptions. The raw perceptions are again taken in, sifted through and integrated, and round and round we go - a perpetual transfer of information between the subject and object occurs. The results from one iteration feed into the next. This exchange between the subject & object defines the data flow on the most basic level.
You should also note that the subject & object are transcendentally related - the information in the outside world is integrated into and ultimately is what makes up the inner world, etc.. And the pattern of interaction between the subject and object as described above is consistent, it does not vary from one person to the next - every person perceives, sifts through the information, stores some of it in memory, remembers things in context, and takes action to effect their environment. This occurs in sequence because each step is a prerequisite for the next - you can't remember something if you haven't stored as a memory yet. So unlike in Model A's incoherent IM model this sequence of steps doesn't vary by type.
data structures & phases
Data structures are configurations of the data at a specific point in the data flow, and are represented by points along the circular continuum in the data flow model. They are discrete, but they divide up the data flow model into phases, which are continuous. The data flow transcends the data structures - it is unidimensional prior to any data structures being imposed onto it. But as we divide up the data flow in this hierarchical fashion, we introduce mechanics through the relationships between IM elements at various levels of the hierarchy. Examples of data structures are the subject and the object, or the intermediary points between the subject and the object. A phase is a period of time inwhich data flows from one structure to another. Phases have a data structure as their input & output, and are represented by the space between data structures in the data flow model. In a phase a transformation in the data occurs. But phase is a broader concept than the traditional concept of functions, because phases exist at every level in the hierarchy, and lower level phases are nested within higher level ones. For example, Ti is nested within the external static phase, which in turn is nested within the broader static phase.
the hierarchy of data structures and phases
To derive the elements we start with the transcendental form - the ring - and we make our first cut through it, forming 2 distinct IM elements. The first two data structures formed we will refer to as the subject & object. A subject & object are the basic actors in the act of information processing. At these points data is at its most subjective and objective forms. You could also say the data is at its most deeply internal or external point here. Data flow between the subject & the object occurs in 2 phases, dynamic and static phases. In the static phase the subject is taking in information - the data originated in the environment, and the subject is taking it in. The data is flowing toward the subject. The subject behaves in a passive manner while taking in information in the sense they are not working to change the environment, hence the term static. The dynamic phase occurs where information emerges from within the subject, and is expressed in the environment. Information is being remembered, or conveyed, or presented, or used to restructure the surroundings. Thus the subject is oriented to exert some influence on the external environment, and so we say this phase is dynamic. You should note that there are some differences in how we define these information aspects from the way Model A defines them, because we're defining them with reference to our data flow model. Refer to the image for a picture of the relationship between the subject / object and the static / dynamic phases. The static functions are Se/Ti/Ne/Fi, whereas the dynamic functions are Ni/Fe/Si/Te.
Next we make a second cut through the ring, perpendicular to the first, forming two intermediary points located directly between the subject and object. These intermediary points are where the subject & object connect and where the transfer of information between the subject & object occurs. The static / dynamic phases are divided up now into 4 phases, and are now additionally characterized by the internal / external aspect. As information flows between the subject & object its state at a given time falls in either side of the intermediary points, relatively closer to either the subject or object. This is the external vs. internal information aspect. The information is either internal (more within the subject) or external (more out there in the world). The external functions are Te/Ti/Se/Si, whereas the internal functions are Ne/Ni/Fe/Fi.
Now we cut the ring again, dividing the existing 4 phases into 8, adding a 3rd level to the IM element hierarchy. The distinction between IM elements at this third level could be characterized in a number of ways. One simple way is by noting which element is relatively closer to the subject vs. object. For example, consider Se and Ti - both external static functions, but Se is in direct contact with the object - it drinks in raw information, whereas Ti is next to the intermediary point, and is where the subject has some more involvement in interpreting and pruning the information. This is the object vs. field distinction, more traditionally known as the introverted vs. extraverted distinction, but better conceptualized in the former way. Field functions are Ti/Si/Ni/Fi, object functions are Te/Se/Fe/Ne.
We have deduced the 8 Jungian functions, along with their common information aspects, from the transcendental form. Refer to the image for a clear picture of what we've done. We're stopping at 8, but if we wanted to continue to 16, 32, and beyond, we could do that.
Notice that the introverted/extraverted function distinction, a traditional and important Jungian concept, is less important here than other aspects - being at a lower level in the hierarchy. It is more significant that Ne and Ni occur within the subject and Se / Si occur outside of the subject. Ne and Ni are both processes of the internal mind, they entail imagination and primary subjectivity. With Si there is an element of subjectivity compared with Se or Te - it is a field aspect - but a person is still not imagining their bodily sensations, those are tangible and relatively objective phenomenon. This is not a process occurring primarily within the persons imagination. The overemphasis on the introverted / extraverted distinction is rooted in long standing Jungian tradition and is unfortunately the source of alot of confusion. For example, functional introversion / extraverted should not be equivocated with a subjects general level introversion or extraversion. Infact, we are not attempting to model introversion / extraversion in our primary model of cognition, we refer only to the object / field aspect and use the language of Jungian functions mainly just to conform with tradition.
Repeating this process once more will result in 16 phases, which will be this models replacement for types. We will discuss the mechanics of this, and the aspects it introduces, later on when we discuss the types in detail.
Descriptions of the 8 functions & their place in the data flow
Fi & the inner self
Notice the point where the subject is located in the data flow (see the image). This point represents the individuals core, innermost self. It's a collection of memories, and it is deeply personal. The information there is persistent and has been organized. Fi is the phase that leads into the innermost point in the data flow (see the image). Fi is responsible for integrating information into the inner self, retaining it as a memory. As the subject integrates information they must organize it - they pick and choose which information to hang on to and which to discard. The information must be brought into a state of consistency with the other information in memory - this is done by reconciling any contradictory parts (the parts that don’t feel good) and retaining the consistent parts (the parts that feel good). Contrary to what many believe about Fi, this is infact a rational process which assesses the coherence or incoherence of information in its broadest context within the subject. New or changed information may trigger a reassessing and restructuring of old memories. Part of reconciling information is stripping it down to its metaphysical form - information is made compatible, and the individuals personal morals are established. This often involves ruminating on the information. Stripping information down also abstracts it, pruning off the unimportant information and hanging onto the meaningful parts. All of this is done with Fi. After Fi has done its job, and the information is done being integrated, it is stored long term and available for recall. Going back to the image, you can see that Fi is pulling information into the subject - this is characteristic of all static functions. It is also dealing with information in the inner world - it is an internal function, i.e. its phase is located on the subjects side of the intermediary divide.
In short, Fi has a functional purpose to organize & maintain the memories.
Ni, associations & recall
Ni is the phase situated directly after the subjective point in the data flow model, where data is first emerging out of the subject, thus Ni is responsible for recollection, ideas, and idealism. Memories live within the subject, where they share a common context and are freely related to one another. Brainstorming and associative recall can occur readily. Memories can be recalled willfully, or the individual can enter an open state of consciousness where they emerge spontaneously. It may not be immediately obvious how the information is relevant, because it's come from within the subject. Memories recollected are often abstract, lacking worldly context, and often come out as vague impressions in an unpresentable form. Thus Ni has been associated with the archetypal imagery of the unconscious. This is not a full characterization though, and can be misleading, because conscious recall is also a Ni process. The critical distinction is that the information is first emerging out of the subject. The ability to freely associate enables creative self-expression and leads to personal revelations - pieces of information connect with seemingly distant and unrelated information, new ideas are apprehended. However, the abstract nature of the information makes it impractical and in need of refinement. The information needs to be found use for; recontextualized with respect to how and where it should fit into the world. Ni doesn’t follow through with practicalities, but it does ponder them; whether an idea is plausible determines whether it's sound. Implausible ideas must be discarded. Still Ni remains mostly caught up within the self, in the creative impulse, the pure inspiration and the flight of ideas. Ni is also associated with ones own ideals - abstract principles held by the subject, and how they should be manifest. This is because Ni is an expression of the morals and axioms established by Fi. Ni is again an internal function (like Fi), and again these are processes that occur within the subject. But it is a dynamic function, and you can see the difference - information (i.e. memories) are emerging out of the subject and are finding their expression.
Ni has a functional purpose to access & bring to bear relevant memories.
To be continued
Reply With Quote
Originally Posted by Eudaemon
