[Date Prev][Date Next][Thread Prev][Thread Next]   [Date Index] [Thread Index] [Author Index] 

Re: Dynamical Hierarchies

Hi David,

You're right. In particular, the statement ("A complete synthetic framework for dynamical hierarchies will not only provide an understanding for the organization and origin of the complexity in biological systems, but also influence all fields that have adopted biological theories or appeal to some form of emergence to create complexity out of simplicty.") is not entirely compatible with Robert Rosen's development of how causality works in a complex universe-- his view was that this IS a complex universe. Therefore, the property of "simplicity" is something that only exists "locally" and is as context dependent as everything else. Another way of putting it is: Simplicity is relational, too.

In order for their statement to work within my father's framework, they would have to qualify certain aspects of what they are saying-- and they might have to change certain other aspects. It's a little hard to tell because there are several possible interpretations that I can see, like the one about their definition of "dynamical". The other points that I think require qualification: There is no way to "create complexity out of simplicity" in general. The only way there can be complex systems in this universe is for the universe itself to be complex. Obviously, if there is no underlying potential (entailment) for complexity, there won't be any! It has to be entailed if it already exists. A complex universe does not forbid "simplicity" the way a simple universe would complexity. As for creating complexity from simplicity in a given, specific, material system, the only way it can be done is via new organization, or reorganization. Even then, the word "organization" has to refer to more than just the material structure. In other words, complexity of any given system is a relational matter and cannot be accomplished by simply adding more parts. One thing I think my father proved rather well is that the behavior of complex systems is not merely a consequence of magnitude of material detail or intricacy.

The aspects of the above statement that I think need to be changed have to do with describing complexity as a "biological theory" or only pertaining to situations where emergence is the process by which complexity happens... That's backwards. Complexity is the set of principles explaining why emergence can happen. Emergence refers to new observable behaviors, or patterns of behavior, spontaneously appearing. As such, complexity is not only a biological theory. Granted, Robert Rosen was a biologist and was trying to answer questions in biology, so it could be said that biological examples provided the breakthrough, but the set of principles is applicable to all systems-- that was the big surprise for him. In fact, if he is correct in his definition of what complexity means and if he is also correct in his assertion that atoms are complex systems then, among other things, there are aspects of those systems which will not be accessible via contemporary physics. I think relational complexity theory is a set of principles which are going to prove to be universally applicable.

Judith
Web address: http://www.rosen-enterprises.com
BioTheory: An electronic journal of general science based on the Relational (Rosennean) Complexity Paradigm
On Oct 8, 2005, at 10:51 AM, David Macy wrote:

<x-tad-smaller>Hey Judith,</x-tad-smaller>
 
<x-tad-smaller>    Yeah, I thought that the introduction was pretty interesting too.  I feel like most of the guys at Artificial Life (AL) are what I call builder/establishers, so if it's a living dynamic (prosaiclly speaking) that they wish to establish and understand, then my guess is that they will arrive at a similar if not identical place that Robert Rosen arrived at.</x-tad-smaller>
 
<x-tad-smaller>It is unclear to me from the context of the introduction what is intended by their use of "dynamical".  That is, whether it is dynamics in state spaces or dynamics in the more prosaic sense.  AL has a "big tent" policy, so maybe they meant it both ways.  Later in the introduction however there is the following sentence...</x-tad-smaller>
 
<x-tad-smaller>"A complete synthetic framework for dynamical hierarchies will not only provide an understanding for the organization and origin of the complexity in biological systems, but also influence all fields that have adopted biological theories or appeal to some form of emergence to create complexity out of simplicty."</x-tad-smaller>
 
<x-tad-smaller>...which strikes me as needing some rewording to be in agreement with Robert Rosen's perspective or usage.  Do you think I'm wrong?  Also my guess is that "a formal framework for synthesizing dynamical hierarchies at all scales" is in fact (M,R)-systems.  I mean don't these "succesive triplets" that Robert spoke of address this issue of dynamical hierarchies at all scales?</x-tad-smaller>
 
<x-tad-smaller>What confounds all these communications I think is a simultaneous effort to define terminology (like dynamical hierarchies) while at the same time using words like complexity, organization, simplicity, emergence, dynamical, etc. which may have a different semantic referents for each who uses them.  What's that old chinese proverb?  </x-tad-smaller><x-tad-smaller>"A word is a finger that points to the moon.  The finger is not itself the moon."</x-tad-smaller><x-tad-smaller>  So you've got the same finger pointing at different moons simultaneously.  Emergence, for instance, I mainly understand as the non-living (organization) to living (organization) transition.  But emergence for some apparently means something like the active site of a protein/enzyme emerging from it's folded shape.</x-tad-smaller>
 
<x-tad-smaller>What's clear is that there are a circle of ideas here that are interelated in some specific manner.</x-tad-smaller>
 
<x-tad-smaller>Oh, apparently Steen Rassmusen is a defender or advocate of what he also calls dynamical hierarchies.</x-tad-smaller>
 
<x-tad-smaller>David</x-tad-smaller>
 <unknown.jpg>
 
 
<x-tad-smaller>----- Original Message -----</x-tad-smaller>
<x-tad-smaller>From:</x-tad-smaller><x-tad-smaller> </x-tad-smaller><x-tad-smaller>Judith Rosen</x-tad-smaller><x-tad-smaller> </x-tad-smaller>
<x-tad-smaller>To:</x-tad-smaller><x-tad-smaller> </x-tad-smaller><x-tad-smaller>***</x-tad-smaller><x-tad-smaller> </x-tad-smaller>
<x-tad-smaller>Sent:</x-tad-smaller><x-tad-smaller> Friday, October 07, 2005 11:32 AM</x-tad-smaller>
<x-tad-smaller>Subject:</x-tad-smaller><x-tad-smaller> Re: Dynamical Hierarchies</x-tad-smaller>

Hi David,

Interesting post!

I wonder how they mean it when they use the word "dynamical"? In my father's usage, dynamical has a particular, very specific meaning. Dynamical systems are simple systems; fully computable (reducible to syntax). They can be addressed via physics, using state descriptions, etc. The more prosaic meaning for "dynamical" tends to mean "constantly interactive and interacting, and therefore undergoing constant change" which is a completely different animal. That usage, and the further description in the clip David posted, seem more to be describing complex system behavior.

But I think the situation is even more difficult than they apparently perceive. In the quote:

"A typical example in this context is the functional differences between proteins and their building blocks, amino acids.  The functionality of proteins is not directly the result of the properties of the individual amino acids. "

The functionality (of proteins) is equally not the result of the entire protein, either. The protein is not capable of functional activity until it assumes its folded shape. Even then, it is further true that the protein's functionality is not just because of the folded shape. Instead, the protein's functionality is a result of the interactions between the folded shape and other aspects of the organism and environment. So all of the previous development of the protein is part of the context for this eventual interaction... Stated another way: The future interactivity (functional capability) of the protein is what drives all of the previous development. That's anticipation, which must be part of the model they seek to build.

But they have achieved a lot here. They clearly perceive that relations between components are of critical interest in modeling organismal behavior, as well as the further fact that direct relations are not the only critical ones in organisms. This represents some much needed, really important progress in science.

Do they, however, recognize the significance of it, with regards to the machine metaphor and the accompanying state-based model of the universe?

Judith


Web address: http://www.rosen-enterprises.com
BioTheory: An electronic journal of general science based on the Relational (Rosennean) Complexity Paradigm
On Oct 5, 2005, at 11:20 AM, David Macy wrote:

 
    I have here in front of me a copy of Artificial Life, volume 11, number 4.  The issue is devoted to dynamical hierarchies.  I thought that I would type in a part of the guest editors' introduction to see what (if anything) you guys might have to say about it.  I generally like these guys at Artificial Life.  They are an eclectic and very cross-disciplinary crowd.
 
The guest editors are as follows:  Tom Lenearts, Dominique Chu, and Richard Watson.
 
The first paragraph of the introduction is as follows...
 
In "Open Problems in Artificial Life", Artificial Life, volume 6 (number 4), pgs. 363-376, Mark Bedau et al. proposed a set of fourteen open problems in artificial life.  The content of this special issue specifically addresses one of those suggested problems: How can we create a formal framework for synthesizing dynamical hierarchies at all scales?  The dynamical hierarchy concept refers to a system that consists of multiple levels of organization having dynamics within and between the entities described at each of the differing levels.  An important aspect of this concept is the fact that entities at different levels can have different functionalities that emerge from the interactions of the lower-level units.  In other words, dynamical hierarchies define a system that is structured by part-whole relationships between objects, where each whole can exhibit properties and can interact in ways different from its parts.   As a consequence, the complete system needs to be modeled as structures relating different description levels of dynamical systems and their interconnectedness.  A typical example in this context is the functional differences between proteins and their building blocks, amino acids.  The functionality of proteins is not directly the result of the properties of the individual amino acids.  The overall protein structure plays a crucial role here.  As a consequence, properties that amino acids do not exhibit in their solitary state can be exhibited collectively.  The same observations can be made when moving from solitary proteins to the level of protein-protein interactions.  New functionalities emerge as a result of the different complexes produced by these interactions.  It is the combination of all these dynamical levels from amino acids to multicellular organisms that makes it a dynamical hierarchy and not merely some simple single-level emergent phenomenon.
 
Well there it is.  I hope that you all are doing well.
 
David