The Causality Gap Between the Non-overlapping Magisteria of Social Pseudoengineering and Post-Modern Pseudointellectualism

Causality in the social pseudosciences is either inadequately principled as practiced by social pseudoengineers or is utterly denied in principle by the post-modern pseudointellectuals.

Much of my focus on Algorithmic Information approximation as model selection in the natural sciences is to get the social pseudoscientists to up their game so that social pseudoengineers have remotely reasonable causal macrosocial models. This isn’t because I’m ignorant of their hidden agendas – but because their agendas are hidden even from themselves by the conflation of is with ought. Nor am I ignorant of the fact that once they plug what they think is their agenda into a remotely reasonable model, that many of them will simply abandon all pretense of rationality – and admit even to themselves that they want to destroy all that anyone might consider good. To borrow a term from Catholic Exorcists, this will expose the “perfection” of their demonic possession. However, there are bound to be some, however few in number, in sufficient possession of their souls to think it possible “in the bowels of Christ” they may have been wrong!

Thus does The Causality Gap close at least a tiny bit from the Alpha side.

But what of the post-modernist pseudointellectuals whose denial of causality is to over-inflate the degree to which causality is an inadequate map of Reality at large? Again, we have a spectrum of possession between those that are merely seduced into confusion about “causality” and those whose possession is closer to “perfection”. The former will, when offered a principled manner in which to approach paranormal phenomena, find in it an attractive opportunity. The “perfected”, however, will find their smoke screens such as “Feminist Epistemology and Philosophy of Science” threatened and react with heightened urgency in any way they can. Although there are plenty of “off-ramps” toward principled investigation of the paranormal, it does seem that the decline of organized religion is making room for a spiritual renaissance, as well as a vacuum in people’s lives that, for many, is being filled by Mammon or worse.

Thus does The Causality Gap close at least a tiny bit from the Omega side.

Having mapped the territory we’re dealing with in The Causality Gap, I’d like to offer a paper by my late ANPA West colleagues, Pierre Noyes and Tom Etter with whom I established a collegial relationship in the 90s as a result of my career-long interest in, if not obsession with, attempting to discover a principled notion of “time” as the foundation of programming language design. Although it is true that this obsession began with Heinz von Foerster handing me a copy of “The Laws of Form” by G. Spencer Brown, wherein Bertrand Russell said he found a resolution to the Theory of Types in the logical analog to Sqrt(-1) as logical self-referential dynamics, it extended to my role as futures architect of the VIEWTRON experiment in Miami circa 1982 and the difficulties of synchronization in network programming. This ultimately led to my hiring Tom Etter at HP’s “Internet Chapter 2” project, circa 2000, to help resolve that long-standing frustration.

The 1998 paper by Noyes and Etter (actually written mostly by Etter) illustrates why I hired Tom Etter in that roll, and it also illustrates why he was then hired by the Boundary Institute to provide principled investigation of the paranormal, once the HP project was exposed as little more than a means of importing huge numbers of H-1bs prior to the DotCon bubble bursting (something that deprived Tom and I of resources to pursue this):

Process, System, Causality, and Quantum Mechanics, A Psychoanalysis of Animal Faith

We shall argue in this paper that a central piece of modern physics does not really belong to physics at all but to elementary probability theory. Given a joint probability distribution J on a set of random variables containing x and y, define a link between x and y to be the condition x=y on J. Define the state D of a link x=y as the joint probability distribution matrix on x and y without the link. The two core laws of quantum mechanics are the Born probability rule, and the unitary dynamical law whose best known form is the Schrodinger’s equation. Von Neumann formulated these two laws in the language of Hilbert space as prob(P) = trace(PD) and D’T = TD respectively, where P is a projection, D and D’ are (von Neumann) density matrices, and T is a unitary transformation. We’ll see that if we regard link states as density matrices, the algebraic forms of these two core laws occur as completely general theorems about links. When we extend probability theory by allowing cases to count negatively, we find that the Hilbert space framework of quantum mechanics proper emerges from the assumption that all D’s are symmetrical in rows and columns. On the other hand, Markovian systems emerge when we assume that one of every linked variable pair has a uniform probability distribution. By representing quantum and Markovian structure in this way, we see clearly both how they differ, and also how they can coexist in natural harmony with each other, as they must in quantum measurement, which we’ll examine in some detail. Looking beyond quantum mechanics, we see how both structures have their special places in a much larger continuum of formal systems that we have yet to look for in nature.

PS: Here is an archived website of related “link theory” papers.

3 Likes

A brief description of how I ended up looking for someone like Tom Etter:

The Mojo language is just another symptom of the same disease that’s been with us ever since Thornton’s superscalar processor for the Cyber 6600 attempted to, in effect, abstract higher level semantics for parallelism from lower level descriptions.

This has been going on long enough folks. Don’t you think?

I mean why torment humans into translating their formulas into state transitions just so that you can torment guys like Thornton and Lattner into imposing enormous complexity into their circuits and/or compilers in an effort to try and divine what non-sequential formulas the humans meant in the first place and reconstitute their inherent parallelism? And by formulas I mean, like what FORTRAN’s inventor, John Backus, meant when his Turing Lecture addressed the need for Formal Functional Programming so that you didn’t prematurely translate your formulas into sequential operations.

I took Backus to lunch at the ACM SIGPLAN '82 conference on lisp functional programming and asked him about relational programming – which everyone seemed to be overlooking. Functions, are after all, degenerate relations.

He hadn’t thought about it.

OK, so where do we go from here?

Although Bruce MacLennan had done a little work at the Naval Postgraduate Research School it wasn’t until I ran across Tom Etter’s link theory that I found someone who had not only approached the problem from the standpoint of relations, but had an approach that looked fundamental enough to describe state transitions in terms of complex numbers. The fact that this was also seen by many as the way forward in specifying quantum computing didn’t impress me since I was far more interested in indeterminacy in network synchronization. David P. Reed tried to deal with that in his PhD thesis. I implemented it in a business setting in the 90s before Reed got it implemented with Alan Kay in the Croquet system. But Reed’s system wasn’t really the right solution since it was functional rather than relational.

It’s really amazing to me that Etter’s extended probability theory has not yet refactored QM, since this is a really obvious thing to do once you accept that negative case counts are just as reasonable as negative bricks of gold.

If people think I’m pissed off at the H-1b fraud for depriving me personally of prestige or money or girls or whatever, they have no idea what I’m really pissed off about. Damn the British for imposing English on their empire.

1 Like

Of non-trivial relevance here is the ongoing cost of the failure to address the metaphysics of dimensional analysis in the foundations of mathematics hence computer programming. Those of us like John Walker who have taken seriously the problem of engineering computation always end up focusing on this fault in the foundations of programming.

The loss of a Mars Orbiter may be one of the more prominent examples but the loss of value on a daily basis to this fault is undoubtedly one of the greatest hidden costs to the age of Moore’s Law and beyond.

PS: As Natural units frequently come up in this context I wonder what would happen if the Planck mass were a basis rather than say the electron mass.

One problem with using the Planck mass as a fundamental unit is that its definition in terms of fundamental constants:

m_{\rm P} = \sqrt{\frac{\hbar c}{G}}

involves the gravitational constant, G (“big G”), which, due to the weakness of gravitation, is known only to around 5 significant digits (22 parts per million). The electron mass, by contrast, is known to around 11 significant digits. This would make the fundamental unit of mass much less accurate than any of the other fundamental units.

In SI, the kilogram is now defined in terms of the second, speed of light, and Planck constant, all of which are known very precisely or defined to be exact values.

Physicists, especially those working in general relativity, sometimes use geometric units in which c and G are set to 1, eliminating messy constants in their equations. This can be further extended by setting Boltzmann’s constant and Planck’s constant to 1, but the quantities one obtains from this do not usually come up in general relativity.

1 Like

That’s an excellent reason to avoid it. I thought of it not just because it is one of the only plausible grand unification constants, but because it appears in the tantalizingly accurate numerology of the Combinatorial Hierarchy, encompassing not only 137 (EM coupling) but gravitational dimensionless coupling, that motivated the creation of ANPA, which is how I came to know both Pierre Noyes and Tom Etter:

1 Like