Endings & Beginnings, or: Continuing the Discussion

[Jerome Warren]

Dear all,

It was fascinating poring over Loet's recent opus, dissecting and inspecting minutae in such a group of gurus and diversely talented scholars.

In lieu of leaving the discussion at the end point of last week's meeting, I thought I would contribute to some further discourse by asking some questions to group members.

@Lucio: I'd be curious to read more about what you think of "triadic reductionism". Also, do you feel that the notion of cyclical redundancy is at all amenable to our notion of "consciousness" or "self-consciousness"? I am thinking in particular of the concept of the "strange loop", which Douglas Hofstedter has used to describe human intelligence, using the symbol of MC Escher's "Penrose Steps". It would be excellent if you had a text to read on these concepts!

@Mark: You mentioned that the distinction I pointed out, between frequentist and "Beyesian" statistics, is very present in your field of computer science. Would you consider wasting some more words describing why you feel this is the case, and whether the driving factor behind that shift is the general move towards "Big Data", i.e., allowing machine intelligence to process huge amounts of data without an explicit underlying model in mind, or whether it is part of a larger epistemological shift away from an asymptotic notion of "truth as recurrence in uniformity"?

Thanks for your contributions,

Jerome

[Ulanowicz, Robert]

Dear Jerome,

I'm not fluent with triadic reductionism, but I attack cycles full-on.
Long ago, I developed an algorithm to identify all the simple cycles
in a network, weigh each one and subtract the cycles from the system.
One is left with a decompositon of the network into two component
networks -- one that is acyclic and a second that consists of pure
cycling. One drawback of the method is that it can get bogged down
when combinatorics become exponential. See:

<https://people.clas.ufl.edu/ulan/files/Cyclng83.pdf>

The reviewers were astounded as to why anyone would be interested in
cycles? Lucio sets them straight! I concur whole-heartedly!

As for consciousness, I see the centripetality of autocatalytic cycles
as defining a virtual vortex of activity with a center. Because all
action circles in towards the center, I see it as a virtual self, id
or ego, characteristic of consciousness. See:

<https://people.clas.ufl.edu/ulan/files/Conscious.pdf>

BTW, the differential equations in the second paper were inserted at
the insistence of the editor, who demands a mechanical cause for
everything. I think they were unnecessary.

Wonderful to be in touch with so many like-minded individuals!

Peace,
Bob U.

P.S. I still owe Lucio a response on these issues. I'm all backed-up
trying to submit a monograph manuscript.

On 8/22/22, Jerome Warren <gregor...@riseup.net> wrote:
>
> @Lucio: I'd be curious to read more about what you think of "triadic
> reductionism". Also, do you feel that the notion of cyclical redundancy is
> at all amenable to our notion of "consciousness" or "self-consciousness"? I
> am thinking in particular of the concept of the "strange loop

> <https://en.wikipedia.org/wiki/Strange_loop>", which Douglas Hofstedter has

> used to describe human intelligence, using the symbol of MC Escher's
> "Penrose Steps". It would be excellent if you had a text to read on these
> concepts!
>

[Lucio Biggiero]

Dear Jerome and all,

thanks for your suggestions. Here I try to address to the issue of triadic redundancy.

Standard (traditional) reductionism is based on finding and dealing with elementary units: from persons, to cells, to molecules, to atoms, etc. Triadic reductionism - which is a label I created ad hoc - is based on triads, whatever the nodes (elements) of a given network are, be them individuals, organizations, cells, etc. Now, let us suppose for a moment that it is always possible to decompose a given network into triads, exactly as it is supposed (by reductionists) possible to decompose a network (or a system) into its nodes (elements). This is all but new, and anti-reductionists (as me and most of you, I guess) know it from ever, but they contend that this possibility does not say much about: 1) a network/system aggregate properties; 2) its dynamics; 3) its determinants. Further, as soon as a network is a bit large and dense, let say 10,000 nodes and 1% normalized density, the number of triads becomes astronomic, preventing (or at least making computationally very hard) any treatment.

However, some of the world leaders in network analysis overlook all previous objections. They promote advanced statistical approaches to network analysis, like the ERGM (Exponential Random Graph Models), and claim that these methods can succeed in understanding a network generation (its determinants) and predicting its dynamics. On the contrary, I argue that it is possible to move to these methods the same criticisms holding for standard reductionism, and especially the most fundamental one: the possibility to decompose in elementary units does not guarantee that, given a set of triads you can recompose it, even if you knew the distribution of triads among the 16 possible types. A fortiori, you cannot either understand its emergent properties or its dynamics. In short: decomposition does not guarantee recomposition.

Further, it's time to question the previous supposition, that it is always possible to decompose a system. In my speech of July 7th 2021 in the broad CoR group, I attacked precisely this point, which in fact is a pillar of any kind of reductionism, including triadic reductionism. I argued that the impossibility of a system decomposability is related to the presence of ... nonlinear functions, which in networks means ... cycles! Exactly the issue on which I'm insisting. If you try to decompose a network structured with cycles, then your decomposition will be very lacking, incomplete, which means that you cannot properly understand it. And the more cycles there are, the more lacking the analysis will be. Moreover, approximation does not hold when a phenomenon is nonlinear, as the studies on deterministic chaos and recursive systems dynamics show very well.

Despite all the arguments against and regardless how solid they can be, scientists never stop running after the chimera of reductionism and full knowledge! That chimera resurges continuingly under different forms. Triadic reductionism is just the last camouflage.

As for the other issue - the connection between redundancy and consciousness - it would deserve a specific meeting, but the provisional answer is yes: in my view, consciousness and meaning are both emergent properties of large-scale recursive networks, as the one that in human beings corresponds to the PNEI (Psycho-Neuro-Endocryn-Immune) System.

Best

Lucio

Lucio Biggiero
Full professor of Organization Science; University of L'Aquila; www.univaq.it
Via Giuseppe Mezzanotte - 67100 L'Aquila (Italy);  lucio.b...@univaq.it; www.luciobiggiero.com  

Department of Industrial Engineering, Information and Economics; tel. +39 0862 434861

Cirps, www.cirps.it; AIRS, www.airs.it; IASS, www.scienzasostenibilita.org 

PEC: lucio.biggiero@legalmail.it; skype: bertagordon; mobile: (+39)3473672426

--
Visit group at https://groups.google.com/g/cogitata?hl=en
---
You received this message because you are subscribed to the Google Groups "Evolutionary Dynamics of Discursive Knowledge" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cogitata+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cogitata/ae128d3d-f102-4735-a463-85e588951fadn%40googlegroups.com.