Facing the Meta-Abstracted Dragon

Jungian Personality Archetypes, Cognitive Synergy, and the Concept of Evil Refracted Through the Prism of Open-Ended Intelligence

Digging a bit into modern Jungian psychology w/ the book "Facing the Dragon" by Robert L. Moore; it's good fun though the style (and the ways in which it's rigorous vs. loose) are very different from the sorts of psychology I'm usually involved in from an AGI perspective.

In pleasantly rabbit-holy fashion, Moore’s thinking led me into a bunch of related musing about the connection between Jungian archetypes and ideas from cognitive systems theory (cognitive synergy, Embodied Communication Prior etc.) … the connection between the notion of Evil and the proper homeostatic functioning of Open-Ended Intelligences … and some other IMO reasonably amusing stuff ;-)

Moore's model of core human personality (presented in his better known book "King, Warrior, Magician, Lover") is based on a handful of Jungian archetypes

• King/Queen, who just sits and radiates glorious energy, inviting fortunate others to resonate with said energy

• Magician, who solves problems via obscure knowledge and clever thinking

• Warrior, who plans out and enacts concerted action

• Lover, who bathes in the wondrous sensations provided by the world, and seeks out experiences of shared lovely sensation with others

Part of the theory here is that a healthy self involved balanced activity and mutually beneficial interaction between the subselves associated with these archetypal characters.

In "Facing the Dragon", Moore focuses centrally on "grandiosity", which he views as a primary mode of action of Evil. This of course is an old idea rendered most beautifully in Goethe's Faust -- Evil, the primal force of separation and destruction, acts on human minds in large part via appealing to the tendency to the human mind to think itself greater and better than it is. By appealing to a person's grandiosity, and offering them the opportunity to (apparently) be more than they actually are, Satan or other avatars of evil convince a person to abandon the good things they actually possess.

Moore reviews the forms that grandiosity takes in personalities dominated by one or another archetype -- e.g. Magician-dominated personalities will tend to get into overly-detached "above it all" fake-enlightenment states (I've fallen into that more than a few times myself), whereas Warrior-dominated personalities will tend to get imperious and abusive, and Lover-dominated personalities will sometimes become grandiosely despondent (threatening suicide and wailing "all is lost" being in itself a way of magnifying the importance of one's own bad feelings beyond the appropriate level).

The solution to grandiosity is posed as simply "knowing one's limits and acting accordingly" .... This is hard to argue with, yet as an entrepreneur and would-be scientific/philosophical revolutionary and general crazy person, I read this and immediately started thinking about how many great things in history would never have been achieved if the folks involved had accepted their limits as commonly conceived. So often real breakthroughs or works of genius -- or simple amazing moments in everyday life -- are achieved by ignoring what one's limits are conceived to be.

As John Lilly said “In the province of the mind, what one believes to be true, either is true or becomes true within certain limits. These limits are to be found experientially and experimentally. When the limits are determined, it is found that they are further beliefs to be transcended. In the province of the mind, there are no limits."

The resolution of this apparent dilemma, I think, lies in digging a little deeper into the nature of reasoning. It's important to know one's limits, but also important to understand the limits of one's knowledge about one's limits. Knowledge about one's own limitations should be framed in terms of uncertainties, rationally estimated based on evidence. Does one really know the limit on how fast one can run; does one really know whether one can climb a certain mountain blindfolded, or prove a certain long-unresolved theorem? Often not, often there is uncertainty involved, and one needs to chart a course between the Scylla of blithe narcissism that assumes not only "everything is possible" but "everything is equally possible for me because I'm so awesome and special", and the Charybdis of defeatist conformism which says "I can't do this because the social accepted 'wisdom' is I can't do this, and all those people can't possibly be wrong."

Part of the subtlety with this is that human minds are not so great at probability estimation, and part of it is that probability estimates are based on assumptions -- assumed knowledge-bases, and more abstract prior assumptions like simplicity measures (if one assumes a la Occam's razor that simplest hypotheses are preferred, a sort of assumption that's needed to cut through the otherwise ungroundability of inductive inference, then one is left with a foundational problem of which measure of simplicity to adopt).

So -- yeah -- one should know one's limits. But one should know them uncertainly, often with quite wide confidence intervals; and one should also know the assumptions on which one's uncertainty estimates are based.

This is a lot for a human mind to handle. And this of course is a large part of the reason why the evil of grandiosity, and the evil of conformist defeatism, both have such an easy time sinking their claws into human minds.

Along the way Moore casually makes remarks suggesting that his core Jungian archetypes correspond to core aspects of the human mind as identified in cognitive psychology, e.g.

• King/Queen = executive decision-making

• Magician = reasoning / deep thinking / creativity

• Warrior = action / planning

• Lover = sensation

(Note that when I refer to "Reasoning" here I'm not focusing just on deductive logic, but including e.g. adductive reasoning which has more of an intuitive or poetic feel to it. The distinguishing factor is that it involves abstraction beyond the immediate, based on long-term memory of diverse particulars that are then converged into novel abstractions.)

This mapping, which he does not make as explicitly as I've done here, seems quite interesting to me and perhaps gives a way of viewing these Jungian ideas about *human* psychology as specializations-to-the-human of more general principles.

If one has ANY generally intelligent agent that's dealing with controlling bodies in a shared environment with other similar bodies controlled by other minds one can expect it to involve cognitive modules corresponding to

• executive decision-making

• declarative, procedural and episodic reasoning

• action and planning

• sensory data processing and reaction to this data

(Because as I've argued in my 2009 paper on the Embodied Communication Prior, these modules correspond to communication modes that are critical in any embodied agent concerned heavily with interacting with other similar embodied agents in a shared environment.)

One then concludes that such an agent is likely to emerge

• a subself that focuses the bulk of its attention on executive decision-making

• a subself that focuses the bulk of its attention on reasoning and deep thinking and creativity

• a subself that focuses the bulk of its attention on action and planning

• a subself that focuses the bulk of its attention on sensation

This will happen because of the same processing constraints that lead humans to develop various subselves -- working memory can only handle a limited amount of information, and there's often efficiency in loading a bunch of closely clustered-together stuff into working-memory, which leads to states-of-consciousness dominated by one or another cognitive module ... and these states-of-consciousness are then meta-cognized leading to sub-(self-models) which then dynamically enact as subselves.

(The dynamics of subselves was covered a fair bit in my 1997 book "From Complexity to Creativity")

The King/Queen, Magician, Warrior and Lover archetypes then appear as the specifically-human versions of these cognitive-module-focused subselves.

And achieving properly reflectively-contextual, higher-order-ly uncertain knowledge of one's limitations requires cognitive synergy among the cognitive control mechanisms associated with one's various memory stores -- which then also requires healthy and respectful multi-party interactions among the various subselves corresponding to these cognitive modules.

The "cognitive synergy" principle underlying much of my work on AGI theory, is thus fairly directly. morphic to the Jungian notion that a healthy human self should have respectively cooperating sub personalities corresponding to the core human-self archetypes. If all this is working wonderfully enough, then one will have a mind that "knows its limits" provisionally in a John Lilly sense, working with them rationally and imaginatively while open to (and spending some resources on) learning new things that overturn one's previous inferences about one's limitations.

(One could cash this out formally in a category-theoretic style by creating a natural transformation mapping the category of cognitive modules into the category of subselves, and then looking at the Jungian personality archetypes as resulting from a projection of the category of subselves into the human sphere.)

In terms of individuation vs. self-transcendence, the key meta-dynamics of mind according to Open Ended Intelligence theory, overly rigid assumption of limits is commonly associated with over-reified individuation and avoidance of radical growth, whereas over-ignorance or willful ignorance of one's limits is often associated with self-destructive behaviors that have the temporary appearance of short-cuts to self-transcendence but actually just leave the system weakened or aborted.

Moore urges us to take Evil seriously as a major player in the universe, and argues that one of the major strengths of Islam and evangelical Christianity in today's world is precisely that they do acknowledge the struggle against Evil as a major feature of human life. Sticking with the Open-Ended Intelligence theme, one is driven to think rather in terms of forces like

• Self-Stagnation -- the opposite of Self-Transformation, the drive to persist the self as it is — minimizing changes

• Self-Explosion aka Grandiosity -- the drive to self-transform faster, bigger and better than a real life-rhythm can support

• Self-Destruction -- the suicide of the individual, not via self-transformation but via choices that lead to simple dissolution of the individual

• Self-Protection -- reification of the boundaries of the individual, so that growth is possible but only if it doesn’t threaten the distinctness and strength of the individual

Basically, in other words,

• insufficient push on self-transformation

• excessive push on self-transformation

• insufficient push on individuation

• excessive push on individuation

Keeping the drives to individuation and self-transformation within the right homeostatic parameter ranges is healthy open-endedly-intelligent functionality, aka "Good" ... letting these drives drift outside the appropriately functional ranges is bad for system survival and/or growth aka "Evil."

This view of Evil as "imbalance" feels very Taoist, at least in an intuitive sense, though I don't have time right now to dig into Taoist literature enough to draw out the relationship with the depth it deserves.

What pushes a system toward Evil is often some sort of struggle with other systems, in a context involving limited resources. It's the limitation of resources, at least in the realms of being we humans typically occupy, that makes the balancing of individuation and self-transcendence into such a difficult optimization or satisfaction problem, thus leading to the parameter-out-of-bounds errors that human theologies classify as Evil. (So this becomes one more variation on the theme that the limitation of resources is the root of all Evil -- as well as the root of all intelligence-as-we-know it, and all artistic-beauty-as-we-know-it, etc. etc. This is why John Lilly followed up the quote I've given above with the observation that "The body imposes definite limits." The domain of the body is the domain of resource limitations. Which in a sense is a construct of the mind anyway -- but the exploration of this construct and its limits is one more thing to be carried out within the domain of the adventurous mind.)

And in the context of specifically human psychology, healthy interoperation of some form of the King/Queen, Magician, Warrior and Lover archetypes corresponds to functional cognitive synergy that makes good enough probability calculations to keep the pushes for individual and self-transcendence within functional operational bounds -- I.e. keeping all Good in the Kingdom/Queendom of the Human Mind.