Manifesto for General Systems Transdisciplinarity

[David Rousseau]

At the ISSS 2015 in Berlin, Germany, we launched our "Manifesto for General Systems Transdisciplinarity" (GSTD), which advocates the need for a General Systems Theory (GST*) and an accompanying General Systems Worldview (GSW) (see attached brochure).  The need for and value of a GST* and GSW, to SE and more widely, was discussed in the SSWG meeting at IW'15, in the INCOSE webinar 76, and two full-day workshops organised by the SSWG and respectively held at the IS'15 in Seattle (July)  and the ISSS 2015 in Berlin (August).  The webinar and workshops focused on how Systems Philosophy can support SE by facilitating the development of general scientific theories in the systems domain, and hence the development and establishment of Systems Science as an academic field.  Systems Philosophy is not itself a science but rather is a branch of the philosophy of science, and as such it can contribute usefully to the advancement of Systems Science.    

The core message of the Manifesto is that Systems Science needs a general theory that can fulfil the role (in the words of Kenneth Boulding) of a unifying “gestalt” for theory building and scientific exploration in the arena of “Systems”  just as the theories of Newton, Lyell, Mendeleev, and Darwin did for Mechanics, Geology, Chemistry and Biology.  We call for renewed efforts towards establishing such a general theory, as originally envisaged by Ludwig von Bertalanffy, one of the founders of the "general systems movement" and the Society for General Systems Research (today the ISSS).  To achieve this general theory will require the input of four distinct groups, namely systems philosophers, systems scientists, and scientists and philosophers more generally.  We value these diverse but distinct perspectives, and call on scientists and scholars to help us develop and establish this much-needed theory and the transdiscipline it would enable. We believe that conditions are now favourable for this to be a practical prospect.

The full Manifesto can be read at http://systemology.org/manifesto.html, where you can also sign it to indicate your support.  Please join us in this important endeavour.

David Rousseau, Jennifer Wilby, Julie Billingham, Stefan Blachfellner     

[Steve Wallis]

Significant progress may be made in a short time by using Integrative Propositional Analysis (the only methodology for rigorously and objectively combining theoretical perspective). It is within our reach to establish a shared space online using a platform such as Kumu or Insight Maker to collaboratively map and integrate our many perspectives. For more localized efforts, I am available as a Fulbright Specialist to visit your institution and facilitate a rapid, effective, and measurable progress.
swa...@projectfast.org

[John Kineman]

My google groups account was apparently set to "no email" so my apologies for not following the list for so long. I think this happened as a result of splitting my personal and Univeristy accounts, which Google forced.  I'm still not sure how it works, but my membership on syssciwg is via my nexial address: jjki...@nexial.org, so it didn't take the response from the university address. I'll try to get it set for both, if that's possible.

John

Begin forwarded message:

From: John Jay Kineman <kin...@colorado.edu>

Subject: Re: [SysSciWG] Manifesto for General Systems Transdisciplinarity

Date: August 15, 2015 at 9:43:56 AM GMT+5:30

To: David Rousseau <david.r...@systemsphilosophy.org>, <syss...@googlegroups.com>

Cc: Judith Rosen <judit...@earthlink.net>, Jennifer Wilby <enqui...@gmail.com>

Thanks David,  I was in the SysSciWG list but I'm not getting the messages now. Maybe because of some account changes with Google? I'll try to re-join.

I fully endorse this effort and I hope we can have a focused sub-theme around it in 2016, perhaps combining the interests of several SIGs and WGs. My plan for 2016 is to do normal SIG and Conference calls for papers, although with some prior guidance on conference themes to those writing the calls. Then the presentation of papers will be in thematic categories based on the abstracts submitted. That way, different SIG ideas on a similar topic can be discussed across SIGs, EGs, and WGs. It seems that would be a good setup for the GST topic, which is gaining popularity across SIGs.  The idea is to encourage certain conference themes and sub-themes in the Calls, then if the membership responds to those calls we should have a response that is at least aimed at the intended topics. If they don't, then that's just the way it is with the membership, but I think if we write the calls properly people will respond accordingly. In the end, the size of each sub-theme discussion would be determined by the number of actual abstracts submitted on that topic, and the topic categories themselves might change from the Call depending on what's actually submitted. I expect there may be quite a number of contributions on the GST topic. After the meeting each SIG, EG, or WG can do their own integration of what was presented/learned/etc. and decide what follow-up their group will do. There can be a collective report at the conference giving status and provisional outcomes, if any; also any agreements with policy implications (like the need for funding and priority) can be fed into the Congress statement. Then long-term development would depend on the SIGs according to their interest and interpretation.

In this way I'm also thinking there is less issue of "ownership" of a topic by a SIG, EG, or WG in the conference itself but maximum ownership and benefit to each group pursuing an aspect or interpretation of the question long-term. As examples, our collective voice at the Conference/Congress can help promote the legitimacy of this discussion, help get funding for specific efforts identified in the SIGs, EGs, and WGs, and also encourage Journals to publish papers on the topic, which many are reluctant to do unless it reinforces the current world view.

Regarding the "Manifesto", the word itself initially took me by surprise. Its technical meaning seems appropriate but my first thought was that it is more commonly used in political contexts, so I wondered how that will be perceived generally. Of course there is a political aspect of getting this on the science agenda, and the idea itself is great -- essential. So perhaps it is the right label if it doesn't produce an unnecessary backlash. We need this focus, and approaching it from a general philosophy, aiming toward development of general theory, makes sense to me. Many practicing scientists, especially in the mainstream, disparage philosophy without understanding its rightful role underpinning their own science. By doing so, they are not arguing against philosophy, but against philosophical discussion that might change current philosophy, thus tacitly accepting the status quo. 

Certainly the general philosophy of science itself has to be examined in this process, along with the systems community reaction to what has been the status quo. It could be that both the commonly received view and the systems community reaction have missed the essence of what is needed; that it is going to talk some work on both sides. Ultimately I think both have to rest on a common foundation.

I think you deserve and have our strongest support in this effort, David. Well done, in my opinion.

John

Dr. John J. Kineman

Senior Research Scientist, CIRES

President, International Society for the System Sciences

john.k...@colorado.edu
(Ph) 303-443-7544: (M) 303-718-9469

On Aug 14, 2015, at 9:16 PM, David Rousseau <david.r...@systemsphilosophy.org> wrote:

Hi John,

 

I don’t know if you are on the SSWG discussion list, but just in case not, here is my posting re the Manifesto for GSTD…

 

Cheers,

David

 

From: syss...@googlegroups.com [mailto:syss...@googlegroups.com] On Behalf Of David Rousseau
Sent: 13 August 2015 18:26
To: Sys Sci Discussion List
Subject: [SysSciWG] Manifesto for General Systems Transdisciplinarity

 

At the ISSS 2015 in Berlin, Germany, we launched our "Manifesto for General Systems Transdisciplinarity" (GSTD), which advocates the need for a General Systems Theory (GST*) and an accompanying General Systems Worldview (GSW) (see attached brochure).  The need for and value of a GST* and GSW, to SE and more widely, was discussed in the SSWG meeting at IW'15, in the INCOSE webinar 76, and two full-day workshops organised by the SSWG and respectively held at the IS'15 in Seattle (July)  and the ISSS 2015 in Berlin (August).  The webinar and workshops focused on how Systems Philosophy can support SE by facilitating the development of general scientific theories in the systems domain, and hence the development and establishment of Systems Science as an academic field.  Systems Philosophy is not itself a science but rather is a branch of the philosophy of science, and as such it can contribute usefully to the advancement of Systems Science.     

The core message of the Manifesto is that Systems Science needs a general theory that can fulfil the role (in the words of Kenneth Boulding) of a unifying “gestalt” for theory building and scientific exploration in the arena of “Systems”  just as the theories of Newton, Lyell, Mendeleev, and Darwin did for Mechanics, Geology, Chemistry and Biology.  We call for renewed efforts towards establishing such a general theory, as originally envisaged by Ludwig von Bertalanffy, one of the founders of the "general systems movement" and the Society for General Systems Research (today the ISSS).  To achieve this general theory will require the input of four distinct groups, namely systems philosophers, systems scientists, and scientists and philosophers more generally.  We value these diverse but distinct perspectives, and call on scientists and scholars to help us develop and establish this much-needed theory and the transdiscipline it would enable. We believe that conditions are now favourable for this to be a practical prospect. 

The full Manifesto can be read at http://systemology.org/manifesto.html, where you can also sign it to indicate your support.  Please join us in this important endeavour. 

David Rousseau, Jennifer Wilby, Julie Billingham, Stefan Blachfellner     

-- 
The SysSciWG wiki is at https://sites.google.com/site/syssciwg/. 
 
Contributions to the discussion are licensed by authors under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.
--- 
You received this message because you are subscribed to the Google Groups "Sys Sci Discussion List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to syssciwg+u...@googlegroups.com.
Visit this group at http://groups.google.com/group/syssciwg.
For more options, visit https://groups.google.com/d/optout.

<Manifesto v8 US Letter size.pdf>

[David Rousseau]

Thanks John, for all this.  I like your plan for the integrating the themes across SIGs, I think this has been a frustration at past conferences where papers on related themes were presented in parallel under concurrent SIG sessions, making it difficult for one to fully engage in a thematic interest.

 

Re the use of the term ‘manifesto’, I hope it will not perceived as somehow politicised.  The term (and the device it signifies) has been used in all areas, there is an overview here: https://en.wikipedia.org/wiki/Manifesto .  Manifestos have been influential in science and education, e.g. the unfortunate one by J. B. Watson that triggered the Behaviourist movement, and the very clever one that advocated the development of the Agile software development methodology: http://agilemanifesto.org/  

 

Glad to have your support in all this, John!

Best regards,

David

John

From: syss...@googlegroups.com [mailto:syss...@googlegroups.com] On Behalf Of David Rousseau

-- 

The SysSciWG wiki is at https://sites.google.com/site/syssciwg/. 
 
Contributions to the discussion are licensed by authors under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.
--- 
You received this message because you are subscribed to the Google Groups "Sys Sci Discussion List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to syssciwg+u...@googlegroups.com.
Visit this group at http://groups.google.com/group/syssciwg.
For more options, visit https://groups.google.com/d/optout.

[Jack Ring]

David, 

pls be aware that the ‘clever’ agile software development manifesto does not result in agile systems. It benefits the developers, not the users.

Pls consider Declaration of Intent, Principles and Method. ConOps is the equivalent.

Jack Ring

[Lenard Troncale]

Dear David,

It is good to have an explicit call (manifesto) for GST. But may I remind you and the others that some have been calling for precisely this for three decades and one might even regard the original and edited versions of LvB's work also as an explicit "manifesto." Further, some of these same folks have actually also suggested explicit pathways or methods for achieving that goal as well as actually consistently DOING what they say should be done. It is always welcome to have new folks join the movement, but it is a little irritating for them to suggest that they are doing it for the first time. It is also a little disingenuous to not make use of the full history of those past and continuing attempts. I am grateful that you cite in some of your slides Klir's work and a couple of others, but there are many more and just naming a group without critically discussing their contributions or where they have fallen short does not advance the field. Unless clear history, attempts, criteria evolve, little more than more people saying "yeah that should be done" will be accomplished.

Len

[David Rousseau]

Dear Len,

 

You misrepresent us when you accuse us of suggesting that we are the first to call for a GST and to indicate ways of attaining it.  The second heading in our Manifesto is “A Renewed Vision” exactly because we see our work as rooted in earlier ambitions, which we are trying to revitalise.  In the references included in the online version of our Manifesto we include both LvB’s book and Boulding’s ‘Skeleton’ paper, and other historical publications that advocated such work.   

 

It is true that we did not give a detailed review of past efforts in this direction in our workshop or manifesto, but the purpose of both were to create a renewed vision and renewed hope for the future, not to assess the lack of past progress in detail.  We are of course aware of the history of GSR, and did highlight it by including in the background reading for the workshops the 2010 paper by Drack (ref below).  That gives a pretty good but very concise summary and assessment of the historical development of General Systems Research (GSR); except for the fact that it does not mention any work by you, apparently because your papers are not included in the academic indexes of databases of peer-reviewed journals for the period they surveyed (1970-2006).  A more detailed overview of the development of GSR is already published in the set of three long papers by Pouvreau and Drack which I sent you before our first workshop six weeks ago (refs below).  These three papers are a very good detailed reference (177 pages!), but once again your contribution is not adequately recognised, and between them they mention only one of your papers (the 1978 one on Linkage Propositions).  Given these published histories there was no need, in our view, to dwell much on these historical details in the workshop.  We did include mention of George Klir because we felt it important to highlight his service to GSR via his editorship for more than 40 years of the International Journal of General Systems.  We also mentioned you, to try and redress the lack of mention of your work in the Drack/Pouvreau papers (you have probably more than 40 relevant papers that have fallen through the academic filters).  We also mentioned you because we think your work is a highly relevant and valuable contribution to one of the three main ways in which a disciplinary field can be unified under a general model/theory, and in this area your work is in our view more relevant and significant than anyone else’s work to date by a very long chalk.  We discussed this in the workshop sessions and will do it again in the series of papers on this subject which we will be submitting to peer-reviewed journals in the next few weeks.

 

In our presentations in the workshop and in the keynote at ISSS 2015 we focused on presenting and defending views that a GST* could exist in principle, that it would be fruitful, what form it might take, that there are realistic strategies for discovering/developing it, and that the philosophical and scientific context is now favourable for this.  We think that these arguments will be helpful in generating new momentum in GSR, and we are hopeful that they will in their way help to advance the field, despite not representing the approach you would have preferred us to take. 

 

David

 

·         Drack, M., & Schwarz, G. (2010). Recent Developments in General System Theory. Systems Research and Behavioral Science, 27(6), 601–610.

·         Drack, M., & Pouvreau, D. (2015). On the history of Ludwig von Bertalanffy’s ‘General Systemology’, and on its relationship to cybernetics – part III: convergences and divergences. International Journal of General Systems, 44(5), 523-571.

·         Pouvreau, D. (2014). On the history of Ludwig von Bertalanffy’s ‘general systemology’, and on its relationship to cybernetics - Part II: Contexts and developments of the systemological hermeneutics instigated by von Bertalanffy. International Journal of General Systems, 43(2), 172–245.

·         Pouvreau, D., & Drack, M. (2007). On the history of Ludwig von Bertalanffy’s ‘General Systemology’, and on its relationship to cybernetics, Part 1. International Journal of General Systems, 36(3), 281–337.

[Lenard Troncale]

Dear David,

Thank you for this detailed response to my personal concerns. I hasten to again pronounce my support for this new manifesto and the hopes you and your supporters have for "generating new momentum in GSR, and (hopes)....(to) help to advance the field."

A Manifesto has to be brief and stirring as a call to arms, so I understand its limitations and applaud your scholarship in attaching reference literature. I welcome the renewal of interest in ISSS, and more widely in systems philosophy, in fulfilling its four organizing principles of ISSS as stated in its By-Laws since its inception. I also mentioned that we need a integrator or synthesizer like Mendeleyev or Mendel or Darwin or Maxwell in my past talks and INCOSE Webinars. I am happy to see that it has been popularized as a nutshell way to express the need the field faces. May your call to arms stimulate much more response and contribution than my calls exactly for "transdisciplinarity" to ISSS over the past 40 years.

Len

On Aug 17, 2015, at 4:07 AM, David Rousseau wrote:

Dear Len,

 

You misrepresent us when you accuse us of suggesting that we are the first to call for a GST and to indicate ways of attaining it.  The second heading in our Manifesto is “A Renewed Vision” exactly because we see our work as rooted in earlier ambitions, which we are trying to revitalise.  In the references included in the online version of our Manifesto we include both LvB’s book and Boulding’s ‘Skeleton’ paper, and other historical publications that advocated such work.   

 

It is true that we did not give a detailed review of past efforts in this direction in our workshop or manifesto, but the purpose of both were to create a renewed vision and renewed hope for the future, not to assess the lack of past progress in detail.  We are of course aware of the history of GSR, and did highlight it by including in the background reading for the workshops the 2010 paper by Drack (ref below).  That gives a pretty good but very concise summary and assessment of the historical development of General Systems Research (GSR); except for the fact that it does not mention any work by you, apparently because your papers are not included in the academic indexes of databases of peer-reviewed journals for the period they surveyed (1970-2006).  A more detailed overview of the development of GSR is already published in the set of three long papers by Pouvreau and Drack which I sent you before our first workshop six weeks ago (refs below).  These three papers are a very good detailed reference (177 pages!), but once again your contribution is not adequately recognised, and between them they mention only one of your papers (the 1978 one on Linkage Propositions).  Given these published histories there was no need, in our view, to dwell much on these historical details in the workshop.  We did include mention of George Klir because we felt it important to highlight his service to GSR via his editorship for more than 40 years of the International Journal of General Systems.  We also mentioned you, to try and redress the lack of mention of your work in the Drack/Pouvreau papers (you have probably more than 40 relevant papers that have fallen through the academic filters).  We also mentioned you because we think your work is a highly relevant and valuable contribution to one of the three main ways in which a disciplinary field can be unified under a general model/theory, and in this area your work is in our view more relevant and significant than anyone else’s work to date by a very long chalk.  We discussed this in the workshop sessions and will do it again in the series of papers on this subject which we will be submitting to peer-reviewed journals in the next few weeks.

 

In our presentations in the workshop and in the keynote at ISSS 2015 we focused on presenting and defending views that a GST* could exist in principle, that it would be fruitful, what form it might take, that there are realistic strategies for discovering/developing it, and that the philosophical and scientific context is now favourable for this.  We think that these arguments will be helpful in generating new momentum in GSR, and we are hopeful that they will in their way help to advance the field, despite not representing the approach you would have preferred us to take. 

 

David

 

·         Drack, M., & Schwarz, G. (2010). Recent Developments in General System Theory. Systems Research and Behavioral Science, 27(6), 601–610.

·         Drack, M., & Pouvreau, D. (2015). On the history of Ludwig von Bertalanffy’s ‘General Systemology’, and on its relationship to cybernetics – part III: convergences and divergences. International Journal of General Systems, 44(5), 523-571.

·         Pouvreau, D. (2014). On the history of Ludwig von Bertalanffy’s ‘general systemology’, and on its relationship to cybernetics - Part II: Contexts and developments of the systemological hermeneutics instigated by von Bertalanffy.International Journal of General Systems, 43(2), 172–245.

[Lenard Troncale]

Dear David,

And please do not use the word "accuse" (as that was not the intent or tone of the msg). I was making observations about the past and not being negative as much as being a bit frustrated as I have said these very same things for decades without much substantive response besides a small few joining my efforts to expand the list of isomorphies, their justification and evidence for, and adding the Linkage Propositions. While in support, I hope you and others understand my feelings of annoyance and of need to represent the past efforts. Perhaps this new renewal can learn from the lack of success of the past?

Len

On Aug 17, 2015, at 4:07 AM, David Rousseau wrote:

Dear Len,

 

You misrepresent us when you accuse us of suggesting that we are the first to call for a GST and to indicate ways of attaining it.  The second heading in our Manifesto is “A Renewed Vision” exactly because we see our work as rooted in earlier ambitions, which we are trying to revitalise.  In the references included in the online version of our Manifesto we include both LvB’s book and Boulding’s ‘Skeleton’ paper, and other historical publications that advocated such work.   

 

It is true that we did not give a detailed review of past efforts in this direction in our workshop or manifesto, but the purpose of both were to create a renewed vision and renewed hope for the future, not to assess the lack of past progress in detail.  We are of course aware of the history of GSR, and did highlight it by including in the background reading for the workshops the 2010 paper by Drack (ref below).  That gives a pretty good but very concise summary and assessment of the historical development of General Systems Research (GSR); except for the fact that it does not mention any work by you, apparently because your papers are not included in the academic indexes of databases of peer-reviewed journals for the period they surveyed (1970-2006).  A more detailed overview of the development of GSR is already published in the set of three long papers by Pouvreau and Drack which I sent you before our first workshop six weeks ago (refs below).  These three papers are a very good detailed reference (177 pages!), but once again your contribution is not adequately recognised, and between them they mention only one of your papers (the 1978 one on Linkage Propositions).  Given these published histories there was no need, in our view, to dwell much on these historical details in the workshop.  We did include mention of George Klir because we felt it important to highlight his service to GSR via his editorship for more than 40 years of the International Journal of General Systems.  We also mentioned you, to try and redress the lack of mention of your work in the Drack/Pouvreau papers (you have probably more than 40 relevant papers that have fallen through the academic filters).  We also mentioned you because we think your work is a highly relevant and valuable contribution to one of the three main ways in which a disciplinary field can be unified under a general model/theory, and in this area your work is in our view more relevant and significant than anyone else’s work to date by a very long chalk.  We discussed this in the workshop sessions and will do it again in the series of papers on this subject which we will be submitting to peer-reviewed journals in the next few weeks.

 

In our presentations in the workshop and in the keynote at ISSS 2015 we focused on presenting and defending views that a GST* could exist in principle, that it would be fruitful, what form it might take, that there are realistic strategies for discovering/developing it, and that the philosophical and scientific context is now favourable for this.  We think that these arguments will be helpful in generating new momentum in GSR, and we are hopeful that they will in their way help to advance the field, despite not representing the approach you would have preferred us to take. 

 

David

 

·         Drack, M., & Schwarz, G. (2010). Recent Developments in General System Theory. Systems Research and Behavioral Science, 27(6), 601–610.

·         Drack, M., & Pouvreau, D. (2015). On the history of Ludwig von Bertalanffy’s ‘General Systemology’, and on its relationship to cybernetics – part III: convergences and divergences. International Journal of General Systems, 44(5), 523-571.

·         Pouvreau, D. (2014). On the history of Ludwig von Bertalanffy’s ‘general systemology’, and on its relationship to cybernetics - Part II: Contexts and developments of the systemological hermeneutics instigated by von Bertalanffy.International Journal of General Systems, 43(2), 172–245.

[Janet Singer]

David

I agree with Len that it would be very helpful to specifically state why it is expected that this integration effort will be successful where others have failed.

(The same comment applies to the SSWG white paper calling for development of SE-SS synergies.)

Janet

[David Rousseau]

Janet,

 

Fair enough!  There have been many GST*-conducive developments in the last while, only a small part of which is due to the current GSTD team.  We are writing up a set of papers on this which we’ll be submitting soon, so for now I’ll only identify these developments rather than give full explanations.  Details about his was covered to some extent in our presentation at IW’15 and formed the basis of the SSWG workshops we did at IS’15 and ISSS 2015.  In no particular order, here are some the reasons for thinking GST has a better chance now than at any previous time: 

 

1.       In the early days, GST was pursued by few researchers, who were working as individuals and mostly doing it in their spare time.  Now there is a big upsurge of interest in GST, as witnessed by the presence of dedicated symposia, discussion groups and special journal issues and the re-issue of classic texts.  As importantly, we are starting to see people with different backgrounds working together as teams rather than as individuals (e.g. the GSTD team and Len’s SPT team), and some researchers are able to focus on general systems research as part of their day job.  This is a big shift from how the task has been approached historically.

2.       Until recently, there were very few practical ideas about how to go about discovering GST*.  The GSTD team developed one based on the synergy between GST* and the GS Worldview, and this is already yielding results we’ll be submitting for publication before the end of the year.  Three further practical-looking strategies came out of the IS’15 and ISSS 2015 workshops, namely ideas about how to work back from the Systemics, distil systems principles by generalizing fundamental disciplinary principles, and leveraging design principles used in engineering. 

3.       For a long time, sentiments in academia have been against the feasibility and even the desirability of developing a GST*.  In the GSTD work we have shown that the philosophical presuppositions behind the vision for a GST* encompasses a range of moderate realisms about the possibility of knowledge, the nature of the natural world, the power of science, the unity of knowledge, the objective existence of concrete systems, and so on.  These positions had low credibility in latter half of the last century, which was dominated (especially in the social sciences) by other kinds of philosophical positions such as Behaviourism, Constructivism and Postmodernism, which made work towards GST* academically suspect, cutting of funds and researcher interest.  These opposing philosophical views are now in strong decline, and Critical Realism and variants of it are rapidly rising in credibility and sophistication.  Moreover, Metaphysics, which has been suppressed in academia for nearly a century is firmly back on the academic agenda, and this year will see the founding of a Society for the Metaphysics of Science.  Many of the questions being asked in the metaphysics of science community bear directly on the systems paradigm and the concerns of the founders of the general systems movement.

4.       In the early days, only a handful of philosophers engaged strongly with the emerging systems paradigm (maybe six).  Mario Bunge did do substantial work on systems, but he did not then or now believe that a GST* can be developed.  However over the last 15 years there has been a strong upsurge in philosophical interest in systems, and there are now many dozens of philosophers of science working on systems concepts, in multiple communities typically outside the historically active interest groups.  Interestingly, many of these philosophers studied and worked in engineering before they became philosophers of science.  This creates a GST-favourable potential which we have not previously had in the philosophy of science. 

5.       Until recently, the term “GST” was used for very many different meanings.  The GSTD team has been able to sort this out, creating a basis for better organised and more productive work on all the aspects of GST, especially what we call GST*, the theory that encompasses the principles behind the evolution and expression of systemic structures and behaviours.

6.       Until recently, the best idea about what GST* might look like was the sketch in Boulding’s 1956 paper “GST – The Skeleton of Science”.  However, based on the work of Julie Billingham over the last few years we now have a more detailed and clearer metaphor to guide us.  We can now see that GST* would be a set of interlocking models and theories that provide a general architecture of a system, principles for instantiating elemental system archetypes, a ‘gestalt’ that reflects the patterns of properties of the archetypes, and principles for combining the archetypes to form complex concrete systems.  This is in the process of being written up for publication.

7.       Until recently, it was unclear on what basis the systems field could be unified.  The GSTD team have identified three routes to unification that represent different kinds of general models/theories, namely models about structures, about processes and about mechanisms.  This enables us to relate the work of different researchers to each other in a constructive way, making it more likely that we can draw effectively on the structure and process aspects as we work towards a general theory about the systemic mechanisms that modulate the systemic processes to produce the systemic structures we find in nature.

8.       Until recently, we had no very clear examples of what general systems principles might look like.  We had only three candidates:  there are no closed systems in nature (Bertalanffy), all viable complex systems are organized as near-decomposable hierarchies (Simon) and the law of requisite variety (Ackoff).  None of these have so far been formulated in a way that would make then stand up as general systems principles, but it looks feasible to do so for at least the first two.  I have discovered two clear general systems principles, which I’m busy writing up for publication.  Based on the example they set I can see how to develop the abovementioned candidates, and I see potential for four more in ideas published by Bunge, Bertalanffy and others.  Once we have a good handful of published general systems principles to work with, general systems research will have wind in its sails.

9.       Much of the work done towards aspects of GST by Len Troncale over the last 40 years is unknown and inaccessible to many who would work on a GST.  That might change in the next year or so, as Len is working on a book bringing it all together and up to date.

10.   Until recently, there was little funding for general systems research and little hope of getting more.  Now things are beginning to look up, with fundraising being planned by the Bertalanffy Centre in Europe, and the NSF is inviting proposals for work to strengthen the theoretical foundations of SE.  The NSF will fund work towards GST if we can show its practical relevance for SE.  I think once there are published general systems principles we will be in a position to develop fundable proposals for improving GST in the interests of SE. 

11.   I have funding in the Centre for Systems Philosophy for a three-year project to develop methodologies that use bits of general systems theory and a simplified version of the general systems worldview to support exploratory science.  I am beginning to see a potential for adapting these investigative methodologies to develop methodologies that support design and engineering.  If this can be done the potential of GST* to support design and engineering will attract good funding and researchers to help improve GST* (this is a different opportunity from the one in #10).   

12.   There is a historical lack of clarity and consistency about the notion of “system”, with different communities using different concepts driven by their local concerns,  e.g. emergent wholeness (biology), human values (sociology) or emergent functionality (engineering).  It is a problem for the vision of a GST* if the “S” in there does not have a stable meaning.  Last month, the ISSS initiated a project to develop a “Systems Science Literacy” guide.  This will have a positive impact for GST studies because this project will have to figure out a representative meaning for “system” before it can say what “systems science” is and why we should study systems.

 

I’ll stop here, to not temp fate by listing 13 reasons for my optimism!

 

David      

[joseph simpson]

David:

Interesting set of ideas.

Just a couple of questions.

What branch of philosophy does systems philosophy identify with and why?

What universals and particulars are addressed in the main area of system ontology or is just one addressed and not the other?

The answer to these questions will help me better understand the current thread of activity.

Take care, be good to yourself and have fun,

Joe

Joe Simpson

“Reasonable people adapt themselves to the world. 

Unreasonable people attempt to adapt the world to themselves. 

All progress, therefore, depends on unreasonable people.”

George Bernard Shaw

[Mike Dee]

Hint:  All

[Me]

Re: (3) and (12)

For those whose views of system are grounded by objectivist metaphysics and the classical theory of categories, and who are open to the findings of cognitive science (a small audience), you may find that the book "Women, Fire, and Dangerous Things", Lakoff, spurs you to reflect on previously unexamined assumptions fundamental to your work.

[joseph simpson]

Steven:

"The Ontological Status Of Women and Abstract Entities," by Alonzo Church is also interesting.

An excerpt may be found here:

http://www.jfsowa.com/ontology/church.htm

Take care and have fun,

Joe

[John Kineman]

Hi David et al.

Sharing a summary and some additional thoughts:

1. GST research is getting more accepted and people are beginning to work in teams. [There are several teams meeting this year about 'Anticipatory Systems Research' based on Rosen's Relational Theory; in Trento, Italy and Capetown SA]

2. GSTD is outlining the strategies for discovering GST

3. "Moderate realism" is being justified to balance the past trend of heuristic construction. [Relational Theory is a quasi-realist approach, meaning it looks at an underlying unity in the causality of systems]

4. Many more people are working on GST, many from engineering [notable work may exist in ecosystem science and certainly in ecosystem management as well, although it is still somewhat 'underground' with respect to the mainstream in the USA. Frameworks such as DPSIR have significant acceptance, especially in Europe]

5. GSTD believes it has developed a more productive approach to discovering GST research by better defining the territory and various strategies. 

6. Based on the work of Billingham, building on Boulding, GSTD believes that "GST* would be a set of interlocking models and theories that provide a general architecture of a system, principles for instantiating elemental system archetypes, a ‘gestalt’ that reflects the patterns of properties of the archetypes, and principles for combining the archetypes to form complex concrete systems." [RT suggests four basic model types in the relational architecture]

7.  GSTD approaches unification of systems theory in terms of structures, processes and mechanisms informing each other. [RT might suggest adding potentials]

8. Until recently there were three commonly accepted candidates for general principles in GST research: (1) no closed systems in nature (Bertalanffy), (2) near-decomposable hierarchies (Simon), and (3) the law of requisite variety (Ackoff). the GSTD team believes the first two can yield some general principle(s). Four more may be possible from other work. We can build a list of GS principles. ["Modeling Relation" might be an important contribution from Rosen]

9. STP needs to be better known after 40 years of development - a book is in the works. [Likewise with other approaches that may be little known because of mainstream oversight; for example IFSR is producing a book on SR based on Action Research frameworks, there are new systemic cosmologies that have a great deal of trouble getting published, a systemic principle of 'Ascendency' (Ulanowicz) or Corning's Synergism principle -- my guess is a lot will come out of the woodwork when it become a generally acceptable discussion]]

10. Funding may be increasing. The US National Science Foundation is funding Systems Research in Engineering and the Bertalanffy Center has funding. Having a set of published principles will help. [Having a set of clear examples that work would really help too]

11. David has funding from the Center for Systems Philosophy to work on "Exploratory Science" that will support design and engineering. [I've recently reviewed experience in the 70's with what came to be known as "Crisis Science" - it has tremendous potential for supporting systems research, integration with sustainability research, and for bootstrapping funding because of its highly relevant application domain - I think it deserves some serious attention. US Dept. of Interior has revived the idea after 30 years, in a new Crisis Science Team]

12. There may be progress in defining what we mean by 'system': ISSS is working on a "System Science Literacy Guide". [I had a discussion with Peter Tuddenham on this yesterday - I recommend referring to it as "Systems Literacy" - we should want it to be about the natural referent itself, which the science is about. The literacy part comes from current science and other knowledge]

John

John Kineman

jjki...@nexial.org

[John Kineman]

Hi David et al.

Sharing a summary and some additional thoughts:

1. GST research is getting more accepted and people are beginning to work in teams. [There are several teams meeting this year about 'Anticipatory Systems Research' based on Rosen's Relational Theory; in Trento, Italy and Capetown SA]

2. GSTD is outlining the strategies for discovering GST

3. "Moderate realism" is being justified to balance the past trend of heuristic construction. [Relational Theory is a quasi-realist approach, meaning it looks at an underlying unity in the causality of systems]

4. Many more people are working on GST, many from engineering [notable work may exist in ecosystem science and certainly in ecosystem management as well, although it is still somewhat 'underground' with respect to the mainstream in the USA. Frameworks such as DPSIR have significant acceptance, especially in Europe]

5. GSTD believes it has developed a more productive approach to discovering GST research by better defining the territory and various strategies. 

6. Based on the work of Billingham, building on Boulding, GSTD believes that "GST* would be a set of interlocking models and theories that provide a general architecture of a system, principles for instantiating elemental system archetypes, a ‘gestalt’ that reflects the patterns of properties of the archetypes, and principles for combining the archetypes to form complex concrete systems." [RT suggests four basic model types in the relational architecture]

7.  GSTD approaches unification of systems theory in terms of structures, processes and mechanisms informing each other. [RT might suggest adding potentials]

8. Until recently there were three commonly accepted candidates for general principles in GST research: (1) no closed systems in nature (Bertalanffy), (2) near-decomposable hierarchies (Simon), and (3) the law of requisite variety (Ackoff). the GSTD team believes the first two can yield some general principle(s). Four more may be possible from other work. We can build a list of GS principles. ["Modeling Relation" might be an important contribution from Rosen]

9. STP needs to be better known after 40 years of development - a book is in the works. [Likewise with other approaches that may be little known because of mainstream oversight; for example IFSR is producing a book on SR based on Action Research frameworks, there are new systemic cosmologies that have a great deal of trouble getting published, a systemic principle of 'Ascendency' (Ulanowicz) or Corning's Synergism principle -- my guess is a lot will come out of the woodwork when it become a generally acceptable discussion]]

10. Funding may be increasing. The US National Science Foundation is funding Systems Research in Engineering and the Bertalanffy Center has funding. Having a set of published principles will help. [Having a set of clear examples that work would really help too]

11. David has funding from the Center for Systems Philosophy to work on "Exploratory Science" that will support design and engineering. [I've recently reviewed experience in the 70's with what came to be known as "Crisis Science" - it has tremendous potential for supporting systems research, integration with sustainability research, and for bootstrapping funding because of its highly relevant application domain - I think it deserves some serious attention. US Dept. of Interior has revived the idea after 30 years, in a new Crisis Science Team]

12. There may be progress in defining what we mean by 'system': ISSS is working on a "System Science Literacy Guide". [I had a discussion with Peter Tuddenham on this yesterday - I recommend referring to it as "Systems Literacy" - we should want it to be about the natural referent itself, which the science is about. The literacy part comes from current science and other knowledge]

John

On Aug 19, 2015, at 7:38 PM, David Rousseau <david.r...@systemsphilosophy.org> wrote:

John Kineman

jjki...@nexial.org

[David Rousseau]

Joe,

 

Systems Philosophy is a branch of the philosophy of science.  From its founding its core intent was to explicate the meaning of the scientific results of the systems scientists, and develop and articulate a worldview that is grounded in the findings of the diverse sciences but informed by the structure and principles of what we now call GST*.  The term “philosophy” in “systems philosophy” has two meanings beyond designating a branch of the disciplinary field of philosophy.  First it refers to doing philosophy about systems (e.g. clarifying concepts etc), and second to being a philosophy in the sense of a worldview, in this case a worldview informed by (general) systems theory.  In the GSTD framework we refer to this latter meaning by the term “General Systems Worldview” (GSW) rather than “the systems philosophy”, so as to try to avoid this ambiguity.

 

Re systems ontology, there are two aspects, one relating to GST* (which is part of science) and the other relating to the GSW (which is part of philosophy).  In regard to GST* general systems researchers are looking for universal patterns and principles that would allow the systemic aspects of particular kinds of systems to be analysed in a consistent way. In regard to GSW, systems philosophers are looking into the particular foundational findings of the specialised sciences and interpreting them as instantiations of the patterns and principles in GST*, and try to construct a worldview that includes a cosmology, a broad description of our best scientific understanding of the nature, organisation, history and potential of the universe.  Of course there are many rival scientific conceptions that might individually qualify as a “best” understanding, so part of this process is to try and select between these rivals on the basis of the need for the ‘big picture’ being developed to be internally coherent, consistent, elegant etc.  For systems philosophers consistency with GST* is one of the ‘virtue criteria’ but there are many others that are part of standard philosophy of science.  An interesting upshot of this approach is that it enables the development of a very compact cosmology (I’m working on this as a reference for the methodology for exploratory science that I’m developing).  

[David Rousseau]

Hi John, thank you for this.  Yes, you are right to point out your work, and the potential of Rosen’s, for the development of a GSTD.  There were many more items I could have listed in my “optimism list” but I thought 12 were enough for a start!

 

I think the Rosen modelling relationship is very important, although it possibly did not originate with him.  There is something very like it discussed in Peter’s Caws’s incoming presidential address to the ISSS back in 1966, nearly 20 years before Rosen’s book.  Peter illustrates his argument by using arguments given by Galileo for his approach.  Introducing Peter’s pre-vision of the Rosen relationship, he says: “Each theoretical system confronts the physical system of which it is the theory, and this confrontation is not a bad image of the human activity we call science.”  Cool, yes?

 

Anyway, I think Rosen’s framing of this relationship is particularly elegant, and of course he developed important philosophical implications from it, e.g. his argument that the modelling relationship belongs to neither the phenomenal nor the formal domain, so its existence is evidence that the idea of reducing all explanations to Physics is hopeless. 

 

I am using Rosen’s modelling relationship in my own research too, for example to extend the AKG model for the purposes of exploratory science methodologies, where the quality of the consilience between phenomenology and theory is exactly what is in question.

 

I am also very interested in the work that you are doing with Judith in using Rosen theory in conjunction with Aristotle’s four causes to develop Holon Theory.  I agree with you about the importance of Aristotle’s idea of the four causes, and I am also working on reviving interest in this, as part of a project with Mario Bunge, with whom I am working on a new model of systems centred on the concept of emergence.   I think in the light of Mario’s model of a system Aristotle can be seen to have conflated two kinds of causes, so actually there are five (I think you have said something similar, for different reasons).  Mario will probably not buy into “final cause”, but like you I think we cannot escape having to deal with what Aristotle was grasping for.  In this project I’m a bit stuck at the moment, due to questions about the systemic nature of the polar types when it comes to the spectrum of systems (the spectrum given Mario’s model of what is a system):  are fundamental particles (at the one end, if there are such things) and the universe (at the other end of the spectrum) systems?  Mario is very clear (universe, yes, ultimate particles, no), but I keep changing my mind on both counts.  I’m beginning to think there is a fundamental problem with the substance-oriented approach to framing questions about the systemness of the phenomenal world.  Maybe a process view is more appropriate?  Anyway, this is good fun to work on, and I am hopeful that by mid next year we’ll have a theory to compare (or better, integrate) with yours.  The good news for General Systems Research is that there are multiple new theories being developed right now that have the potential to make a fundamental difference to how we think about the category “systems”.  This represents a key shift from how things have stood for the last 40-odd years.

 

David

[joseph simpson]

David:

Thanks for the additional contextual information, it helps.

Take care, be good to yourself and have fun,

Joe

[David Rousseau]

Stephen,

 

Thanks for this.  To clarify, the search for a GST* (a scientific endeavour) is not a search for a grand narrative, but rather merely a search for a theory under which the systemic aspects of the specialised disciplines can be understood in a consistent way. 

 

However, Systems Philosophy (a philosophical endeavour) is looking for a worldview informed by GST*, and this includes a cosmology which is something like a ‘grand narrative’, although this has to be understood in a rather modest way.  I am aware that it is no longer thought to be the case, in mainstream philosophy of science, that there can be “bridging laws” between the specialised disciplines, so we are at best looking for some sort of consilience of knowledge rather that some sort of grand unification grounded in physics or mathematics or consciousness or something else.  More pertinently, we have to acknowledge that even though the universe may be a unity we only encounter limited patches of it, and this in a sort of accidental way, and that this contingency together with the limitations of our science, intelligence and cognitive apparatus means that we are unlikely to be able to ever weave a coherent unified worldview out of our patchy investigations.  But even so, for those of us who are epistemic realists and broad naturalists and believe in the unity of the universe it remains an ideal to work towards some kind of consilience between our fragmentary theories, even though we acknowledge our substantial limitations in pursuing this ambition.

 

In the light of these uncertainties and difficulties I am an advocate of an eclectic approach, and would wish to take multiple views into account, and not be completely constrained by e.g. the limits of the philosophy of science.  I think that the philosophy of SE and the philosophy of technology provide independent valuable insights, as you very rightly also point out.  It was for this reason that we included the papers by Pennock and Wade (philosophy of engineering) and Mitcham (philosophy of technology) in the pre-reading material for the workshop.  That said I would like to learn much more about both these areas, so thank you very much for your guiding comments and the reference.  This may be important in the future of the GSTD team’s research, as ideas came out of the workshops (both in Seattle and Berlin) about how to use the principles of successful design in IT and in engineering to try and discover principles for a GST*. Also, as I pointed out in another post to the SSWG discussion group, I am beginning to see a way to transition my analytic project to a synthetic one, so the shift from a science to design/engineering perspective is very relevant for me.

 

With warm regards,

David          

 

 

From: Stephen Cook [mailto:stephe...@adelaide.edu.au]
Sent: 20 August 2015 07:26
To: david.r...@systemsphilosophy.org
Cc: j.w...@hull.ac.uk
Subject: RE: Manifesto for General Systems Transdisciplinarity

 

David

 

I enjoyed your tutorial at the INCOSE IS and made a number of points to Jennifer that I will repeat in case they have not found their way to you.

 

Firstly, are we really striving to derive a GST* or as Checkland would call it a framework of ideas?  My interpretation of the quest for GST* is equivalent to the search for a Grand Narrative; not seen by many as being a useful pursuit.  In the sciences there are many theories that hold currency simultaneously and in areas like management science it is well recognised that there is not, and perhaps will never be, an overall theory of management.  Useful sets of theories, knowledge, information and data can, however, be bundled into a coherent framework of ideas.

 

Checkland posits that for each area of concern (A) there are recognised methodologies (M) that can be applied each of which draws on its associated Framework of ideas (F).   The question then becomes to what extent can these frameworks be combined into a single framework?  Some would argue that this is not a sound proposition because each comes from a particular tradition eg, functionalist, interpretive, critical, etc.  Such people favour multi-methodological approaches that draw on more than one of these traditions.

 

Finally, on the subject of the philosophy of systems engineering, I would suggest that you don’t start from the history and philosophy of science but rather from the philosophy of engineering or the philosophy of technology.  The latter philosophical schools have a different foci from science; engineering values what works and uses this knowledge to advance the discipline.  Engineers always conduct design work for products and systems without complete knowledge of why things work, they just know they do.  Over time, engineers and others usually seek to provide explanations as to why things work as observed.  A simple example is dental implants.  It is known that bone tissue not only tolerates titanium but actually integrates with it.  This has become the basis of a large industry but why titanium works whereas other metals do not is yet to be understood.  Thus we have engineering knowledge on this subject but scant scientific knowledge.

 

There is a small literature on the philosophy of engineering.  A good place to start is the Royal Academy of Engineering: http://www.raeng.org.uk/policy/engineering-ethics/philosophy  Engineering draws heavily on design (which is synthetic) so the design literature will also contain some gems.  (Science is fundamentally analytic, another reason why the philosophy of science may not be the best starting point.)

 

I hope this has been helpful.

 

 

Professor Stephen Cook, PhD, FIET, FIEAust, INCOSE Fellow

Adjunct Professor

ECIC

The University of Adelaide

 

Level 5

10 Pulteney St

Adelaide, 5005

Australia

 

Tel: +61 (0) 41 882 9946

 

 

 

 

 

 

 

From: David Rousseau [mailto:david.r...@systemsphilosophy.org]
Sent: Friday, 14 August 2015 6:55 PM
To: David Rousseau <david.r...@systemsphilosophy.org>
Subject: Manifesto for General Systems Transdisciplinarity

 

To:  All who registered for the INCOSE SSWG Workshops on “Systems Philosophy and its relevance to Systems Engineering”

 

On behalf of Jennifer, Julie, Stefan and myself I’d like to thank you all for your participation and inputs at the workshops in Seattle and Berlin.  We enjoyed meeting you and learnt much from the experience and from you – thank you!  We hope to see you again at follow-on events.

 

As you will remember, the workshops included a brief mention of our plan to launch a "Manifesto for General Systems Transdisciplinarity" (GSTD) at the ISSS 2015 in Berlin, Germany.  We did this as a plenary address on Tuesday 4 August, calling for renewed efforts towards developing a General Systems Theory (GST*) and an accompanying General Systems Worldview (GSW) (see attached brochure).    

 

We spent some time in the workshops discussing the potential value of a GST* and the GSW for Systems Science in general and SE in particular, and we hope that you will support and participate in the called-for efforts towards establishing and leveraging general systems models and theories.

 

The full Manifesto can be read at http://systemology.org/manifesto.html, where you can also sign it to indicate your support.  Please join us in this important endeavour.

 

David, Jennifer, Julie and Stefan     

 

[John Kineman]

David,

A few minor points..

--  Indeed the modeling relation is also the epistemic cut; I think we can find it throughout many works including von Neuman and others with a different application of it. I've tried to trace it back to the Vedas, so certainly establishing precedence would be problematic. However, I don't think who mentioned it first is important except for the history books, what's more relevant is who gave it mathematical form and made it useful in certain ways. Rosen used it to argue incompleteness of the current view (essentially) while also pointing out that it is indispensable, precisely because it is a picture of science itself, possibly knowledge itself. If his treatment of it leads us to a more complete theory, then he deserves the credit for describing it in the most meaningful way, whereas it is rather obvious in many contexts - it is also the mind-brain relation graphed simply. Another analogy would be Koestler's introduction of the idea of holon (maybe he wasn't first either?). I use the term and reference Koestler for having introduced it with the right philosophical idea, but Koestler didn't give the logic or math of how it works. Another analogy, evolution was a well known concept before Darwin. But Darwin gave a convincing idea of how it might work. I think Rosen's was perhaps the most extensive and deep treatment of modeling relations. But definitely cool!

--  On its implications, yes, that's one interpretation. Another, which I take, is that the fact that the information relations are needed but not part of either natural or formal system means that we need a broader concept to contain modeling relations. If modeling relations are truly general (and thus our showing natural and formal systems isn't some expandable parochialism about what those system concepts are, but already their most expanded concept) then the only candidate for a larger system is self-reference; a holographic modeling relation. This brings information into the study of nature including both natural and formal systems. We of course are asking just that question in relating, for 2016 for example, "human" and "natural" systems. Today's science is begging for the answer on how to couple across the modeling relation. So the alternative to this "impossibility" is to expand physics, which is what Rosen advocated. The "New Physics".

-- On Aristotle, actually I see Aristotle not has having invented this - like your argument for precedents of the modeling relation. He's the one who screwed it up. He basically took what was known from ancient times, and very available to the Greeks at the time, and turned it from a cycle to a hierarchy. Instant duality. Mario might be more comfortable with final cause after understanding how the cycle works and howe it keeps entirely within nature. Final cause in Aristotle's terms is unacceptable to science. But the duality he established (along with others of course) was just the kind of specificity that was needed to progress with mechanistic science. As Rosen pointed out, it limited science to studying this very special kind of system - the mechanism, and we are glad for it. But it thus didn't study the rest of nature, and I think that's what we're getting back to now.

-- Yes, there is a problem with the substance approach and I would agree with process orientation as an alternative. Properties are not about things, they are "thingness" abstracted from a complex system. The property is the only reason we think it is a thing and not a process that generates properties. Not wishing to be too presumptuous, but incapable of behaving differently, I'll venture a response to your dilemma: "are fundamental particles (at the one end, if there are such things) and the universe (at the other end of the spectrum) systems?" I believe the relational view, fully developed, would say everything is a system, except for how we view it in fractions. Only systems can exist in the universe because only systems are (qualified) 'real'. This presumes a definition of system that includes its creation and its operation - as the modeling relation does. It also could be said to agree with the physicists, that ultimately the only 'reality' (that we can assign) is consciousness. And if consciousness isn't a system that spawns systems, it isn't anything at all. But I suppose that's getting too strange for most folks. More practically, RT would say that the particle, as we know it, is known by its measurements. The measurements are abstractions from a more complex reality -- a system. So by definition a particle is a fraction of a system, but ontologically it is but one aspect of a whole that is not a fraction. The fraction cannot exist alone. So, Koestler's insight that holons are both part and whole at the same time. The measurement is always tied to the system being measured and the system doing the measuring (the modeling relation). So, for my money, the modeling relation is the definition of a system.

-- Indeed the discussions should be interesting, and it does seem that the energy is ramping up on this otherwise forgotten topic.

John Kineman

jjki...@nexial.org

[Janet Singer]

John,

In response to David’s point 

12. There may be progress in defining what we mean by 'system': ISSS is working on a "System Science Literacy Guide”. 

you wrote

I had a discussion with Peter Tuddenham on this yesterday - I recommend referring to it as "Systems Literacy" - we should want it to be about the natural referent itself, which the science is about. 

By ‘natural referent’ do you mean something like:

[A] natural system is a set of qualities, to which definite relations can be imputed. As such, then, a natural system from the outset embodies a mental construct (i.e. a relation established by the mind between percepts) which comprises a hypothesis or model pertaining to the organization of the external world. (Robert Rosen, Anticipatory Systems, 1985)

Janet

[John Kineman]

Janet,

Yes, certainly. It is further clarified in the next paragraph:

“In what follows, we shall refer to a perceptible quality of a natural system as an observable. We shall call a relation obtaining between two or more observables belonging to a natural system a linkage between them. We take the viewpoint that the study of natural systems is precisely the specification of the observables belonging to such a system, and a characterization of the manner in which they are linked. Indeed, for us observables are the fundamental units of natural systems, just as percepts are the fundamental units of experience."

So he distinguishes observables of a natural system from percepts of a formal system. Observables are traditionally "experience of the outer world" - these are percepts that anyone can confirm, i.e. part of the general entanglement of humans with their environment; whereas inner percepts without outer confirmation may not be part of the general entanglement that gives us the impression of a persistent reality outside (which of course can't be known by any other means). We learn the difference by testing hypotheses about those observables to see where they are predicated; i.e. in the mind itself, or with something outside. None of this is controversial or non-traditional. He was showing how knowledge is bootstrapped and  he was being extremely precise about where, when, and how so later he could show where a limiting assumption was made that reduced science to the idea of mechanisms

If you recall my holarchical "nesting" of modeling relations within each side of a modeling relation, you then see what is being described when he says that a mental construct is also involved in our image of the natural system. The natural system is also a modeling relation (my words now, not Rosen's explicitly). Making it self-similar modeling relations all the way is of course non-traditional, but otherwise it is just the 'epistemic cut'.

With regard to my comment on literacy, i'm saying that it should be about the observables (natural system) not the percepts (science, etc.). In other words we want to be literate about something imputed to the natural world as "system" in the same way that we would want to be literate about the "ocean" the "climate" the "economy" or even "evolution" as a process - the literacy should be about the phenomena in question, not the study of it as such. They are different levels of mental construct.

I can perhaps anticipate the question, are systems natural phenomena in the same way that oceans are? Is systemicity a perceived organization that we assign to nature but that may not actually be in nature as such? I would say the whole point of introducing the idea of "systems literacy" should be to say it is a natural object of study in its own right, rather than ceding the territory of presumed reality to reductions alone. Process philosophy also comes very close to doing that. Compare it with "evolution literacy", for example. We can be literate about physical objects, processes, patterns (as in landscape ecology) and origins (as in evolution and cosmology). 'System' is an imputed organization of all these things. So, in my view, it would be "Literacy about the organization of nature" - perhaps the grand literacy. 

Dr. John J. Kineman

Senior Research Scientist, CIRES

President, International Society for the System Sciences

john.k...@colorado.edu
(Ph) 303-443-7544: (M) 303-718-9469

[James Martin]

I would like to clarify what I believe Rosen meant when he referred to "natural system":

[A] natural system is a set of qualities, to which definite relations can be imputed. As such, then, a natural system from the outset embodies a mental construct (i.e. a relation established by the mind between percepts) which comprises a hypothesis or model pertaining to the organization of the external world. (Robert Rosen, Anticipatory Systems, 1985)

There are two kinds of so-called natural systems:

1) systems that exist in the world (as opposed to in the mind)

2) systems that occur in the world due to natural forces alone (ie, not man-made)

Biologists, geologists, physicists, astronomers, and the like, are mostly concerned with natural systems of the 2nd kind. My understanding of Rosen is that he was referring to natural systems of the 1st kind in his depiction of the modeling relation, as seen below. Rosen's formal system per this construct is basically our thoughts about a system that exists in the world (or one that might or could exist).

From this I conclude that our engineered systems, when realized in the world (ie, built from our design), are natural systems of the 1st kind (ie, the Rosen kind). So, we need to be careful when we talk about engineered systems vs natural systems that we make it clear we are talking about contrasting what we create with those systems that Nature creates (of the 2nd kind).

Furthermore, our engineered systems, when they exist in our mind, or on paper or some other form of expression, are formal systems in the Rosen sense. From this we can conclude that the "systems" we are engineering are of these two kinds -- the formal system when it exists as a conceptualization or a design, and later as a natural system (of the Rosen kind) when this design is realized.

Therefore, we need to pay attention to the issues that Rosen is raising when systems engineers "think" and talk about the systems we engineer as only of one kind, not the two that Rosen is concerned with. I assert that this conflation of concepts leads to many of our problems when doing systems engineering.

We as engineers have as our goal to conceive of solutions (formal systems in the mind) that address one or more problems by eventually this transforming this formal system into a natural system (realized in the world).

James

James

[Jack Ring]

James, 

I encourage you to read Rosen carefully, particularly his use of the word hypothesis. 

This may cause you to avoid using “exist’ and ‘occur’ and perhaps restate your claims as ‘systems that are hypothesized to exist in the world and the subset that are hypothesized to occur sans man.

This may help elucidate the concepts for which the words may be misleading. 

Hopefully it will ensure the application of the scientific method to vet such hypotheses. 

FWIW, I presume System Science is specifically intended to vet all such hypotheses about (supposed) systems.

FWIW, I recommend pondering what Derek Cabrera calls Perspective in his DSRP construct.

?? was that a tree falling??

Onward,

On Aug 22, 2015, at 5:37 AM, James Martin <mart...@gmail.com> wrote:

I would like to clarify what I believe Rosen meant when he referred to "natural system":

[A] natural system is a set of qualities, to which definite relations can be imputed. As such, then, a natural system from the outset embodies a mental construct (i.e. a relation established by the mind between percepts) which comprises a hypothesis or model pertaining to the organization of the external world. (Robert Rosen, Anticipatory Systems, 1985)

There are two kinds of so-called natural systems:

1) systems that exist in the world (as opposed to in the mind)

2) systems that occur in the world due to natural forces alone (ie, not man-made)

Biologists, geologists, physicists, astronomers, and the like, are mostly concerned with natural systems of the 2nd kind. My understanding of Rosen is that he was referring to natural systems of the 1st kind in his depiction of the modeling relation, as seen below. Rosen's formal system per this construct is basically our thoughts about a system that exists in the world (or one that might or could exist).

<image.png>

James

[Janet Singer]

John, James, Jack,

I agree that consideration of Rosen's modeling relation can resolve most if not all disagreements over the definition of 'system'. (Michael Lissack, President of the American Society of Cybernetics, used Rosen's modeling relation in his own keynote on 'building bridges between interpretations' at ISSS2015 in Berlin a few weeks ago.)

The modeling relation is particularly useful for showing the complementarity between the science and engineering perspectives, as James pointed out. But Jack is also correct in his caution re the need for care in understanding the subtleties of Rosen's approach. Some relevant sections from Tim Gwinn's article The Rosen Modeling Relation

An important point to emphasize is this: a natural system is itself a kind of “hypothesis or model” of the external world. At first glance, this might seem to invoke some kind of circularity in the Modeling Relation, in the sense of having the system under study be itself a model; however, the situation is more subtle than this. In particular, we must distinguish between a natural system and a material system. ...

The third subtlety relates to realization. Realization is the process of working from a formal system to a natural system. [5d] It is essentially the notion of going from a blueprint to a working material version. When realizing a system, the process does not involve every intimate detail of its material structure. Instead, it is sufficient that the resultant material version embody the criteria from the formal model (the “blueprint”). So, the congruence relation established is only between the elements and relations specified in the formal model and a corresponding certain finite number of observables and relations in the material realization. As noted in the second subtlety, those identified observables and relations are abstractions. Whatever other additional material characteristics the realization might have, to the extent that they do not affect the congruence relation, such additional characteristics are not part of the natural system. Therefore, a natural system is not synonymous with a material system. Instead, a natural system is some subjectively defined subset, or abstraction, of an actual material system.

The article is worth reading in its entirety to provide common context for clarifying the 'system definition' question in this forum, at the IW, and/or a Linz conversation.  http://www.panmere.com/?page_id=18

(Judith, did your dad write a similar short summary of his own?)

Janet

[Judith Rosen]

I think James is accurate in his assessment; Robert Rosen defined a natural system as any system that exists in the world (as opposed to in the mind).

A formal system-- for example on paper, such as a blueprint-- can be analyzed as a natural system also but would be a totally different "thing" if we did so. If we looked at it according to its organization as a natural system: it's a piece of paper with information imprinted on it. It's got a material structure we can analyze which as nothing much to do with the ink printed on the surface. We can ask, how did it come into being? It did not self-organize. It was created by human beings for reasons that also come from human beings. Nearly all the entailment for the existence of it comes from outside the system itself. It is actually a subsystem of the human beings who created it. 

That is the beauty of my father's definition for what is a System: "A collection of percepts that seem, to us, to go together in some way."

Consider for a moment what a "definition" is... It cannot exist outside of human language, which cannot exist outside of human thought and interaction, which cannot exist outside of human life and mind... A definition is about meaning. As such, it has human perspective and experience built into it. Furthermore, any collection of percepts we want to point to and call a system is also, simultaneously, a component of countless other systems. At each level, we can do the same analysis of looking at the organization of the system and learn useful things about it. This is why I have advised against trying to nail down some sort of definition regarding what a "generic" system is. The very benefit of systems thinking is to allow us to get away from specific instantiations and begin to understand how organization itself drives causality.

Judith

PS: Sorry if you receive this twice; I wasn't actually subscribed to this and couldn't post this response but didn't realize that until it bounced. I think I've got it fixed now...

[Lenard Troncale]

Dear Judith,Janet, James and John, (sounds like the bible)

As has happened so many times before in these streams of discussion on syssci, the tangent taken from the original Manifesto stream has diverged onto a more particular stream on perhaps "Definition of System." I do not know how to change these "subject" titles post facto, but it would help us all if, when a stream topic changes, it would be changed in the title stream to help electronic storage, filing, recover, searches etc. later. This is VERY important as the msg number gets into the thousands.....

On SysSciWG there are also many other streams of msgs on "Definition of System" that these should be gathered into. Same goes for the very large msg exchanges on NECSI by scientists. I have over 9,000 or those (on many varied topics) far exceeding SSWG to date.

As for the Definition of System.....   I welcome reference to what past authorities have written, but we need to formulate our consensus in our time period, learning from them and possibly extending. This field is way too fluid and new to simply cite past authorities in my humble opinion. Past authorities may have pointed the way, but much has been learned and digested over these decades that must be incorporated and considered.

In my integration, natural systems would include those studied by the sciences, especially those that existed stably and sustainably without humans or human thought about them for all but the last million years (virtually all of astronomy, cosmology, aspects of physics, chemistry, geology, all but human biology, etc.). But inescapably, humans build real systems too. And humans have natural origins. So we would be hard pressed to call them "artificial" systems as past authorities like Herbert Simon tended to do. If they require human invention and intersession to remain in existence, then we might want to make a distinction, but not necessarily natural or not. Human engineered and sustained entities are still systems. But perhaps they have less of the isomorphies required for natural systems to be sustained.

I wish we could come to a consensus on this endless debate and issue at the foundation of the field or movement or manifesto.

Len

[John Kineman]

One thing that might help in this regard is to change a title by indicating the former title with a colon to the changed, or sub-title, like a parent-child relation. Often my dilemma in changing a title is that it will no longer be seen as part of a stream that generated it. That is, change of topic is not always that abrupt that we should divorce it from its origin. So, I tried this in the current title - we'll see if its feasible. From this transition point, future posts could be titled "suggestion on threading discussions", since the history of the change is preserved here.  

Just a thought.

John

John Kineman

jjki...@nexial.org

[Jack Ring]

Perhaps we could examine several systems that have a) satisfactory effects on their contexts, b) when and while needed, and c) are considered trustworthy by the users, then determine how they do that then determine how they got to be that way. 

[David Rousseau]

Len,

 

I think we can make a useful distinction by being careful with how we apply the terms “natural” and “naturalistic”.   

 

It seems to me intuitive to reserve the term “natural system” for concrete systems we find in nature, and use some other term like “artifactual system” or “engineered system” for concrete systems we make out of nature (and use some other term like “conceptual system” for the abstract systems we conceive of).  

 

Natural and artifactual systems both qualify as naturalistic systems, because their behaviours and properties are in accordance with the laws of nature.

 

In terms of this framing, to designate something as a “naturalistic system” is to signal something about the kinds of processes it can support, whereas to designate something as a “natural system” or an “artifactual system” is to signal something about the kinds of processes that produced it.

 

David

[James Martin]

David,

You said that natural systems are "concrete systems we find in nature". It seems to me that there are also abstract natural systems, such as the collection of natural laws that govern the behavior of planets in the solar system, for example. These laws interact with each other in such a way that you get emergent behavior that individual laws don't engender. 

Or perhaps a better example is the feedback process that many natural systems incorporate into their architecture. This process is an abstract system unto itself for maintaining a stable condition. It can be thought of as a subsystem in the overall system that uses this process to control some important parameter. 

James

Sent from my iPad

[John Kineman]

I agree with what Len said, specificallyl:

As for the Definition of System.....   This field is way too fluid and new to simply cite past authorities.....

In my integration, natural systems would include those studied by the sciences, especially those that existed stably and sustainably without humans or human thought about them ... But inescapably, humans build real systems too. And humans have natural origins.... we might want to make a distinction, but not necessarily natural or not. Human engineered and sustained entities are still systems. 

But I think also that Rosen's definition of a system that Judith quoted is not at all inconsistent with this sentiment. For reasons of developing relational theory he began with a general description of what 'system' means recognizing foremost the epistemic problem in knowing anything about what actually 'is' in nature. 

Then also James's wish to find some distinction between human engineered system and natural ones can be accommodated within both views without proposing that these systems are of a different fundamental organization. The relational theory produces a logical structure in which both 'kinds' of realization of a system can be explained by the same fundamental organization of all systems, natural or artificial.

Rosen left the synthesis to his followers, but gave tantalizing hints, somewhat like Fermat's last Theorem -- a puzzle that others could solve if they study the work in enough depth. At this point we have a tentative mathematical description in category theory that needs testing and further development. It can easily show the difference between engineered systems and the more random constructions that we call 'natural'. But to see that distinction requires first understanding modeling relations.

There were three or four stages in developing the theory of modeling relations. It may be easiest to consider them in the order of the most obvious to the most problematic for modern science to see why Rosen started with biology but reached the conclusion that the theory would "change physics".

1. The most obvious case was science itself. The modeling relation describes in great detail why there is an 'epistemic cut' and how it works in both gaining and limiting knowledge. Examples here include everything from machines to life and consciousness, but assuming only one modeling relation and one single-formalism as the basic diagram suggests. It applies to the modern science paradigm that presumed, hoped, pursued, and eventually disproved the idea that one formalism of natural law would be enough. The fact that these relations may not 'commute' perfectly means that at least some of nature is not mechanistic, while we do otherwise see (and make) apparently solid objects and mechanisms as quite prevalent realized systems in the general environment. A hint came from atomic research, which found physical incompleteness at supposedly fundamental levels, requiring more than one formalism. At the same time we discovered that formal systems can't be fundamentally complete either. That requires explanation -- why do there appear to be two kinds of system, one that has a mathematically complete model and one that does not? Does completeness arise from incompleteness, or possibly the other way around?

2. Its easiest now to jump to the most obvious case of incompleteness and thus complexity, which is that of internal modeling relations in living systems, cells, organisms, and by their composition, ecosystems. Life itself turns out to be a causal closure within the organism (Rosen's M-R system) and also with the environment (phenotype and genotype). Thus organisms have behavioral and evolutionary models that they encoded themselves, like Maturana and Varela's 'autopoietic' systems. There's no point arguing if these organismic 'models' are 'real' or 'natural'; it is the alternative (that they are only human concepts) that becomes unnecessarily complicated and unworkable; for if they are not internal models then we cannot explain their 'impredicativity' (lack of predication on general causality).

3. The first case involving one modeling relation suggested that such a relation cannot be complete (i.e., must have other models). The case of life involved a special organization of multiple modeling relations. Now we can back up to consider the intermediate case between 1 and 2 above, that of 'merely' complex systems (not living). That case requires two or more causally open modeling relations (full causal closure requires five, which is the living case). Here only an open pair of incompatible (mutually irreducible) models whose realizations are in the same material system will suffice. Examples include quantum phenomena (wave-particle 'duality'), dissipative systems (export of entropy implies increase of internal order, i.e., an internal model), and probably cosmology (as yet confounded on the split between relativity and quantum mechanics). Of course this kind of complexity can also appear as an emergent property of living systems, so a familiar example might be husband and wife each trying to manage the same household. Immiscible rule sets make the interaction complex. The most basic case of dual models, that applies to every system, is the model for existence and the model for operation. It seems these two formal domains are not mutually reducible for anything. That suggests that everything is fundamentally complex; hence the question posed above is answered that completeness, or the impression of it, arises from incompleteness, not the other way around.

4. Finally we can deal with the most problematic case for modern science, that of mechanisms and apparently 'solid' or 'concrete' -- both terms meaning 'well-defined' -- systems. These are the appearances in nature and human manufacture that seem most 'real' to us (recall Rosen's use of that phrase in the definition Janet quoted). They are now trivially explained as reductions of the complex to the first case, where one complete model seems to exist at our level of interaction. The reduction itself is easily understood as the result of many interactions (like the concept of quantum 'decoherence', a term that is itself stated from the opposite view).

So, now I can use the above theory structure to say what the difference is between engineered and 'natural' systems; and how they can both have the same underlying organization. 

I know of nothing in Rosen's writings that restricts application of modeling relations to the human case, although the first example given was of science itself. That's basically the epistemic cut but in sufficient detail to see how science became reductionistic. That starting point was not a limitation, but a necessity because the epistemic case is the only one we know directly. We know that we create models. We thus can infer that other systems to too. Then, testing that empirically, we find considerable confirmation of model-based behavior everywhere, even in the previously thought 'concrete' material foundation of particulate, state-based physics, which instead became probabilistic. There is nothing concrete about it now - it dissolves into conscious relations.

Then Rosen described the next most obvious case of modeling relations in biology - life is characterized by closure of certain internal models. Life creates and employs natural internal models (Rosen's M-R systems as the most elemental case).

Then we can back up to a non-living system that is nevertheless complex. That's an intermediate case between a single modeling relation, as in science, and what turns out to be a closure of five modeling relations in life. Complexity involves two modeling relations involving at least two incompatible formal systems, not necessarily closed to each other except for sharing the same realization. An analogy would be husband and wife both managing the same household, but applying incompatible models. 

, i.e., two incompatible models. Since the models are unknown to each other (not mutually reducible) there is an impredicativity from either side, making the sum complex. Finally, even mechanism can be seen as a modeling relation - a non-complex one. That is, what was previously thought to be the foundation to build up from, defined objects and mechanisms, becomes a special case of the complex, which is general. Its a complete reversal of the paradigm, which is why it is so hard to grasp or promote, why he chose not to try to splash it in the New York Times, and also why he said it would change physics. We can borrow pieces of the work to patch various existing views, but the fact remains that the whole of our thinking has to flip to embrace relational thinking. He said, for example, that the key to the new science is to "objectify the impredicativity" - which is the mis-match between system models. That means to make complexity the natural object of study and treat the material reductions to state and dynamics as the abstractions. Materiality is based on quantitative measures of state, but Rosen wrote: "there can be no greater act of abstraction than the collapsing of a phenomena in N [the natural system] to a single number, the result of a measurement" (Life Itself, pg. 60). In other words, the things we can measure, the observables, previously used to define a natural system, are not what's objective about the natural system. A critical phrase in that definition that Janet provided was "for us". For an observer making models, the natural system must be defined by such abstractions (observables) because that is all we can see of it; but he is saying through mathematical language that the true reality must be considered to be complex relations from which measurements abstract concepts of state on which we then model dynamics. If he says to objectify the impredicativity (and throw away state-based physics), clearly he is saying we must objectify the modeling relation itself as an inferred concept of reality underlying the measurements. We can consider it experienced in the human case as epistemology.

That is an internalized modeling relation. More dramatically he said "throw away the physics [as currently pursued] and keep the organization". Modeling relations describe a supervening organization of a system, inherent to all systems. He also wrote: "There is nothing unphysical in the relational strategy".

Read, for example, from:  Hiley, B. & Peat, F.D. (2012) Quantum Implications: Essays in Honour of David Bohm, Routledge.

(Sorry for the cliff hanger at the end - I don't have this book, just got the preview page in Google Books).

Is it modeling relations all the way? Since the modeling relation is a fundamental picture of 'knowing', in humans and nature, it is impossible to 'know' an exception. The only way there could be an exception is if a natural system could exist without any models .. but now recall the definition in which 'natural system' involves knowledge, thus models. It does not work to say these are only human models - all models are about something that has models. In describing the relational process of science, he was showing that we are really bringing human models (built on percepts) into congruence with natural models (built on observables). Its a masterful piece of reasoning that leaves no logical alternative to saying that all systems have natural models. 

 and models are natural, even human ones. Living systems are more, however, in that they are characterized by closure of multiple models. I've demonstrated that following Rosen's logic these turn out to be five: two inside the organism itself (Metabolism and Repair exactly as he described M-R systems), and three shared with the environment (Replication, Behavior, and Selection - yielding phenotype and genotype).

 It appears to be the rule rather than the exception. But we still haven't made it mathematical. Not yet a technical analysis that can be applied. That's where the second hint comes in - category theory entailments. I'll leave that, because Rosen didn't put the two together - these are the ingredients that I tried to assemble into a synthesis in 2011. Suffice to say that I'm convinced it can be done (maybe is started) and that the result is a mathematical description of a generic system, or rather a new way of analyzing systems in terms of whole systems.

This was where Rosen began with Rashevsky to develop Relational Biology, which led full circle to the idea that it actually gives a new view of the physics -- that it is the special case organizationally (i.e., a reduction) not the general one. Hopefully you can see the magnitude of this conclusion, and the reason he did not try to splash it on the front page of the New York Times. 

However, I also sympathize with James' sentiment that there must be some distinction between engineered systems and those that seem to develop out of supposed 'natural' laws; i.e., not obviously involving mental intent or design. But the distinction is not what modeling relations describe and what they don't. A theory based on modeling relations actually allows us to explain the difference.

As you can see in Level 4 above, modeling relations do not necessarily imply complexity. It could be a fully commuting mechanism where the model, i.e., formal mathematical description that we call a model, is a 'largest model' - in other words it is complete in and of itself. I also think of it as a singular model (unlike in quantum theory where we need two incompatible models, wave and particle). What makes it complex, then, is when there is no largest model - the models are incomplete. Theoretically, then even wave and particle wouldn't do the trick, the combination of those models is also incomplete. But even if wave and particle models could exhaust the territory, it would still mean there is no one single complete largest model because the wave and particle models can't be combined in a single formalism - they are based on different assumptions about nature - different formal causes. While people may look for a synthesis that captures both, Rosen complexity would say there isn't one - there will always be some 'semantic residue'; something that the meaning of a wave and the meaning of a particle does not quite get.

But beside from that further complexity, which can be left for a deeper analysis, the most basic proximal picture of complexity is a modeling relation with two models. That is two modeling relations that apply to the same system. And the most generic two models of anything are its behavior in the world, and its existence in the world. I suppose one could make a case that particle define local existence and waves describe non-local potential for existence (interference patterns, double paths, etc.). In any case, it is obvious that organisms have two internally generated models, one for their ecological behavior and another for their existence, or reproduction. Much of their life is about getting their behavior to support their existence as a system; i.e, their system identity. And in natural living systems these two models run simultaneously so both can change to come into congruence. Evolution may be rapid or slow, seeking this balance or having found it.

Now the case of an engineered system. The last thing the design engineer wants is something that modifies its existence and behavior. The engineer and market decides its existence/design and modifies that according to market success, but we don't what the engineered system to do it by itself. If your car changes into a toaster while you are driving, you would be very disappointed. So a goal of engineering something is to separate its operation from its design. But of course even with a completely physical machine we can't do that completely; all systems will re-design themselves to some degree, which is why there is always a "mean time before failure". But if we include the designer and factory, in fact it does behave like an evolving species. Many companies even take individual failed products back to analyze the failure and correct it - a kind of artificial selection.

In ecology, the boundary is very blurry now. It used to be preserved, but no longer with the idea of "niche defining phenotypes" (Odling-Smee), which is now recognized. In other words, many organisms (if not all to some extent) engineer their environment. A bird or gorilla nest may differ in sophistication from the Stanley hotel, but with regard to purpose they are more alike than different. There are also organisms that 'borrow' artifacts (e.g., hermit crabs). And there are humans who live very close to the land. So, where's the boundary?

From what I can tell reading Rosen, he was very keenly aware of the epistemic problem - that we only know our own experience and there is no direct knowledge of an outside world (also Descartes point in his discourse on method). The modeling relation expresses that idea. What we directly experience are the percepts in our own experiential apparatus. However, I don't think he meant to equivocate on the existence of an outside world. The point in understanding science is what we can know (epistemology), but

[John Kineman]

I accidentally hit the send button instead of save - still composing this to get it readable, so better to wait before getting a headache....

I agree with what Len said, specificallyl:

> As for the Definition of System..... This field is way too fluid and new to simply cite past authorities.....
> In my integration, natural systems would include those studied by the sciences, especially those that existed stably and sustainably without humans or human thought about them ... But inescapably, humans build real systems too. And humans have natural origins.... we might want to make a distinction, but not necessarily natural or not. Human engineered and sustained entities are still systems.

.....etc.

[John Kineman]

OK, here's the edited version. Please ignore the other (I've deleted the trail because I'm responding specifically to what Len wrote below).

I agree with what Len said, specificallyl:

As for the Definition of System.....   This field is way too fluid and new to simply cite past authorities.....

In my integration, natural systems would include those studied by the sciences, especially those that existed stably and sustainably without humans or human thought about them ... But inescapably, humans build real systems too. And humans have natural origins.... we might want to make a distinction, but not necessarily natural or not. Human engineered and sustained entities are still systems. 

But I think also that Rosen's definition of a system that Judith quoted is not at all inconsistent with this sentiment. For reasons of developing relational theory he began with a general description of what 'system' means recognizing foremost the epistemic problem in knowing anything about what actually 'is' in nature. 

Then also James's wish to find some distinction between human engineered system and natural ones can be accommodated within both views without proposing that these systems are of a different fundamental organization. The relational theory produces a logical structure in which both 'kinds' of realization of a system can be explained by the same fundamental organization of all systems, natural or artificial.

Rosen left the synthesis to his followers, but gave tantalizing hints, somewhat like Fermat's last Theorem -- a puzzle that others could solve if they study the work in enough depth. At this point we have a tentative mathematical description in category theory that needs testing and further development. It can easily show the difference between engineered systems and the more random constructions that we call 'natural'. But to see that distinction requires first understanding modeling relations.

There were three or four stages in developing the theory of modeling relations. It may be easiest to consider them in the order of the most obvious to the most problematic for modern science to see why Rosen started with biology but reached the conclusion that the theory would "change physics".

1. The most obvious case was science itself. The modeling relation describes in great detail why there is an 'epistemic cut' and how it works in both gaining and limiting knowledge. Examples here include everything from machines to life and consciousness, but assuming only one modeling relation and one single-formalism as the basic diagram suggests. It applies to the modern science paradigm that presumed, hoped, pursued, and eventually disproved the idea that one formalism of natural law would be enough. The fact that these relations may not 'commute' perfectly means that at least some of nature is not mechanistic, while we do otherwise see (and make) apparently solid objects and mechanisms as quite prevalent realized systems in the general environment. A hint came from atomic research, which found physical incompleteness at supposedly fundamental levels, requiring more than one formalism. At the same time we discovered that formal systems can't be fundamentally complete either. That requires explanation -- why do there appear to be two kinds of system, one that has a mathematically complete model and one that does not? Does completeness arise from incompleteness, or possibly the other way around?

2. Its easiest now to jump to the most obvious case of incompleteness and thus complexity, which is that of internal modeling relations in living systems, cells, organisms, and by their composition, ecosystems. Life itself turns out to be a causal closure within the organism (Rosen's M-R system) and also with the environment (phenotype and genotype). Thus organisms have behavioral and evolutionary models that they encoded themselves, like Maturana and Varela's 'autopoietic' systems. There's no point arguing if these organismic 'models' are 'real' or 'natural'; it is the alternative (that they are only human concepts) that becomes unnecessarily complicated and unworkable; for if they are not internal models then we cannot explain their 'impredicativity' (lack of predication on general causality).

3. The first case involving one modeling relation suggested that such a relation cannot be complete (i.e., must have other models). The case of life involved a special organization of multiple modeling relations. Now we can back up to consider the intermediate case between 1 and 2 above, that of 'merely' complex systems (not living). That case requires two or more causally open modeling relations (full causal closure requires five, which is the living case). Here only an open pair of incompatible (mutually irreducible) models whose realizations are in the same material system will suffice. Examples include quantum phenomena (wave-particle 'duality'), dissipative systems (export of entropy implies increase of internal order, i.e., an internal model), and probably cosmology (as yet confounded on the split between relativity and quantum mechanics). Of course this kind of complexity can also appear as an emergent property of living systems, so a familiar example might be husband and wife each trying to manage the same household. Immiscible rule sets make the interaction complex. The most basic case of dual models, that applies to every system, is the model for existence and the model for operation. It seems these two formal domains are not mutually reducible for anything. That suggests that everything is fundamentally complex; hence the question posed above is answered that completeness, or the impression of it, arises from incompleteness, not the other way around.

4. Finally we can deal with the most problematic case for modern science, that of mechanisms and apparently 'solid' or 'concrete' -- both terms meaning 'well-defined' -- systems. These are the appearances in nature and human manufacture that seem most 'real' to us (recall Rosen's use of that phrase in the definition Janet quoted). They are now trivially explained as reductions of the complex to the first case, where one complete model seems to exist at our level of interaction. The reduction itself is easily understood as the result of many interactions (like the concept of quantum 'decoherence', a term that is itself stated from the opposite view).

So, now I can use the above theory structure to say what the difference is between engineered and 'natural' systems; and how they can both have the same underlying organization. 

The case of an engineered system can be described as the case of separating the existence and behavior models so they don't act simultaneously. The engineer and market decides existence/design like the evolutionary environment of an organism. The last thing we want is for the engineered system to do that by itself. If your car changes into a toaster while you are driving, you would be very disappointed! So a goal of engineering something is to separate its operation from its design. But of course even with a completely physical machine we can't do that completely; all systems will re-design themselves to some degree, which is why there is always a "mean time before failure". So, the 'whole' system includes the designer and factory. Just the car or the toaster are a 'fraction' of the system, the abstracted material state-based machine. Somehow much of physical nature also has a separated common origin in the distant past. I think that occurs because of interaction, like a decoherence phenomenon. Over time, causally open system develop a common existence domain.. a common history.

Cheers,

John

[John Kineman]

Michael,

Yes, I accidentally hit send instead of save -- see the edited version just sent. But if you are interested in some of the other detail, I don't think there is anything contradictory there, just more rambling. I was working toward a concise statement. Take it as evidence that I DO try to be brief, just rarely succeed.

Cheers,

John Kineman

jjki...@nexial.org

On Aug 24, 2015, at 11:25 AM, Michael Singer <mjsi...@soe.ucsc.edu> wrote:

Hi John — was your email to syssciwg accidentally cutoff? It ended in mid-sentence.

Michael

On Aug 23, 2015, at 10:05 PM, John Kineman <jjki...@nexial.org> wrote:

I agree with what Len said, specificallyl:

As for the Definition of System.....   This field is way too fluid and new to simply cite past authorities.....

In my integration, natural systems would include those studied by the sciences, especially those that existed stably and sustainably without humans or human thought about them ... But inescapably, humans build real systems too. And humans have natural origins.... we might want to make a distinction, but not necessarily natural or not. Human engineered and sustained entities are still systems. 

But I think also that Rosen's definition of a system that Judith quoted is not at all inconsistent with this sentiment. For reasons of developing relational theory he began with a general description of what 'system' means recognizing foremost the epistemic problem in knowing anything about what actually 'is' in nature. 

Then also James's wish to find some distinction between human engineered system and natural ones can be accommodated within both views without proposing that these systems are of a different fundamental organization. The relational theory produces a logical structure in which both 'kinds' of realization of a system can be explained by the same fundamental organization of all systems, natural or artificial.

Rosen left the synthesis to his followers, but gave tantalizing hints, somewhat like Fermat's last Theorem -- a puzzle that others could solve if they study the work in enough depth. At this point we have a tentative mathematical description in category theory that needs testing and further development. It can easily show the difference between engineered systems and the more random constructions that we call 'natural'. But to see that distinction requires first understanding modeling relations.

There were three or four stages in developing the theory of modeling relations. It may be easiest to consider them in the order of the most obvious to the most problematic for modern science to see why Rosen started with biology but reached the conclusion that the theory would "change physics".

1. The most obvious case was science itself. The modeling relation describes in great detail why there is an 'epistemic cut' and how it works in both gaining and limiting knowledge. Examples here include everything from machines to life and consciousness, but assuming only one modeling relation and one single-formalism as the basic diagram suggests. It applies to the modern science paradigm that presumed, hoped, pursued, and eventually disproved the idea that one formalism of natural law would be enough. The fact that these relations may not 'commute' perfectly means that at least some of nature is not mechanistic, while we do otherwise see (and make) apparently solid objects and mechanisms as quite prevalent realized systems in the general environment. A hint came from atomic research, which found physical incompleteness at supposedly fundamental levels, requiring more than one formalism. At the same time we discovered that formal systems can't be fundamentally complete either. That requires explanation -- why do there appear to be two kinds of system, one that has a mathematically complete model and one that does not? Does completeness arise from incompleteness, or possibly the other way around?

2. Its easiest now to jump to the most obvious case of incompleteness and thus complexity, which is that of internal modeling relations in living systems, cells, organisms, and by their composition, ecosystems. Life itself turns out to be a causal closure within the organism (Rosen's M-R system) and also with the environment (phenotype and genotype). Thus organisms have behavioral and evolutionary models that they encoded themselves, like Maturana and Varela's 'autopoietic' systems. There's no point arguing if these organismic 'models' are 'real' or 'natural'; it is the alternative (that they are only human concepts) that becomes unnecessarily complicated and unworkable; for if they are not internal models then we cannot explain their 'impredicativity' (lack of predication on general causality).

3. The first case involving one modeling relation suggested that such a relation cannot be complete (i.e., must have other models). The case of life involved a special organization of multiple modeling relations. Now we can back up to consider the intermediate case between 1 and 2 above, that of 'merely' complex systems (not living). That case requires two or more causally open modeling relations (full causal closure requires five, which is the living case). Here only an open pair of incompatible (mutually irreducible) models whose realizations are in the same material system will suffice. Examples include quantum phenomena (wave-particle 'duality'), dissipative systems (export of entropy implies increase of internal order, i.e., an internal model), and probably cosmology (as yet confounded on the split between relativity and quantum mechanics). Of course this kind of complexity can also appear as an emergent property of living systems, so a familiar example might be husband and wife each trying to manage the same household. Immiscible rule sets make the interaction complex. The most basic case of dual models, that applies to every system, is the model for existence and the model for operation. It seems these two formal domains are not mutually reducible for anything. That suggests that everything is fundamentally complex; hence the question posed above is answered that completeness, or the impression of it, arises from incompleteness, not the other way around.

4. Finally we can deal with the most problematic case for modern science, that of mechanisms and apparently 'solid' or 'concrete' -- both terms meaning 'well-defined' -- systems. These are the appearances in nature and human manufacture that seem most 'real' to us (recall Rosen's use of that phrase in the definition Janet quoted). They are now trivially explained as reductions of the complex to the first case, where one complete model seems to exist at our level of interaction. The reduction itself is easily understood as the result of many interactions (like the concept of quantum 'decoherence', a term that is itself stated from the opposite view).

So, now I can use the above theory structure to say what the difference is between engineered and 'natural' systems; and how they can both have the same underlying organization. 

I know of nothing in Rosen's writings that restricts application of modeling relations to the human case, although the first example given was of science itself. That's basically the epistemic cut but in sufficient detail to see how science became reductionistic. That starting point was not a limitation, but a necessity because the epistemic case is the only one we know directly. We know that we create models. We thus can infer that other systems to too. Then, testing that empirically, we find considerable confirmation of model-based behavior everywhere, even in the previously thought 'concrete' material foundation of particulate, state-based physics, which instead became probabilistic. There is nothing concrete about it now - it dissolves into conscious relations.

Then Rosen described the next most obvious case of modeling relations in biology - life is characterized by closure of certain internal models. Life creates and employs natural internal models (Rosen's M-R systems as the most elemental case).

Then we can back up to a non-living system that is nevertheless complex. That's an intermediate case between a single modeling relation, as in science, and what turns out to be a closure of five modeling relations in life. Complexity involves two modeling relations involving at least two incompatible formal systems, not necessarily closed to each other except for sharing the same realization. An analogy would be husband and wife both managing the same household, but applying incompatible models. 

, i.e., two incompatible models. Since the models are unknown to each other (not mutually reducible) there is an impredicativity from either side, making the sum complex. Finally, even mechanism can be seen as a modeling relation - a non-complex one. That is, what was previously thought to be the foundation to build up from, defined objects and mechanisms, becomes a special case of the complex, which is general. Its a complete reversal of the paradigm, which is why it is so hard to grasp or promote, why he chose not to try to splash it in the New York Times, and also why he said it would change physics. We can borrow pieces of the work to patch various existing views, but the fact remains that the whole of our thinking has to flip to embrace relational thinking. He said, for example, that the key to the new science is to "objectify the impredicativity" - which is the mis-match between system models. That means to make complexity the natural object of study and treat the material reductions to state and dynamics as the abstractions. Materiality is based on quantitative measures of state, but Rosen wrote: "there can be no greater act of abstraction than the collapsing of a phenomena in N [the natural system] to a single number, the result of a measurement" (Life Itself, pg. 60). In other words, the things we can measure, the observables, previously used to define a natural system, are not what's objective about the natural system. A critical phrase in that definition that Janet provided was "for us". For an observer making models, the natural system must be defined by such abstractions (observables) because that is all we can see of it; but he is saying through mathematical language that the true reality must be considered to be complex relations from which measurements abstract concepts of state on which we then model dynamics. If he says to objectify the impredicativity (and throw away state-based physics), clearly he is saying we must objectify the modeling relation itself as an inferred concept of reality underlying the measurements. We can consider it experienced in the human case as epistemology.

That is an internalized modeling relation. More dramatically he said "throw away the physics [as currently pursued] and keep the organization". Modeling relations describe a supervening organization of a system, inherent to all systems. He also wrote: "There is nothing unphysical in the relational strategy".

Read, for example, from:  Hiley, B. & Peat, F.D. (2012) Quantum Implications: Essays in Honour of David Bohm, Routledge.

(Sorry for the cliff hanger at the end - I don't have this book, just got the preview page in Google Books).

<Screen Shot 2015-08-23 at 2.15.43 PM.png>

Is it modeling relations all the way? Since the modeling relation is a fundamental picture of 'knowing', in humans and nature, it is impossible to 'know' an exception. The only way there could be an exception is if a natural system could exist without any models .. but now recall the definition in which 'natural system' involves knowledge, thus models. It does not work to say these are only human models - all models are about something that has models. In describing the relational process of science, he was showing that we are really bringing human models (built on percepts) into congruence with natural models (built on observables). Its a masterful piece of reasoning that leaves no logical alternative to saying that all systems have natural models. 

 and models are natural, even human ones. Living systems are more, however, in that they are characterized by closure of multiple models. I've demonstrated that following Rosen's logic these turn out to be five: two inside the organism itself (Metabolism and Repair exactly as he described M-R systems), and three shared with the environment (Replication, Behavior, and Selection - yielding phenotype and genotype).

 It appears to be the rule rather than the exception. But we still haven't made it mathematical. Not yet a technical analysis that can be applied. That's where the second hint comes in - category theory entailments. I'll leave that, because Rosen didn't put the two together - these are the ingredients that I tried to assemble into a synthesis in 2011. Suffice to say that I'm convinced it can be done (maybe is started) and that the result is a mathematical description of a generic system, or rather a new way of analyzing systems in terms of whole systems.

This was where Rosen began with Rashevsky to develop Relational Biology, which led full circle to the idea that it actually gives a new view of the physics -- that it is the special case organizationally (i.e., a reduction) not the general one. Hopefully you can see the magnitude of this conclusion, and the reason he did not try to splash it on the front page of the New York Times. 

However, I also sympathize with James' sentiment that there must be some distinction between engineered systems and those that seem to develop out of supposed 'natural' laws; i.e., not obviously involving mental intent or design. But the distinction is not what modeling relations describe and what they don't. A theory based on modeling relations actually allows us to explain the difference.

As you can see in Level 4 above, modeling relations do not necessarily imply complexity. It could be a fully commuting mechanism where the model, i.e., formal mathematical description that we call a model, is a 'largest model' - in other words it is complete in and of itself. I also think of it as a singular model (unlike in quantum theory where we need two incompatible models, wave and particle). What makes it complex, then, is when there is no largest model - the models are incomplete. Theoretically, then even wave and particle wouldn't do the trick, the combination of those models is also incomplete. But even if wave and particle models could exhaust the territory, it would still mean there is no one single complete largest model because the wave and particle models can't be combined in a single formalism - they are based on different assumptions about nature - different formal causes. While people may look for a synthesis that captures both, Rosen complexity would say there isn't one - there will always be some 'semantic residue'; something that the meaning of a wave and the meaning of a particle does not quite get.

But beside from that further complexity, which can be left for a deeper analysis, the most basic proximal picture of complexity is a modeling relation with two models. That is two modeling relations that apply to the same system. And the most generic two models of anything are its behavior in the world, and its existence in the world. I suppose one could make a case that particle define local existence and waves describe non-local potential for existence (interference patterns, double paths, etc.). In any case, it is obvious that organisms have two internally generated models, one for their ecological behavior and another for their existence, or reproduction. Much of their life is about getting their behavior to support their existence as a system; i.e, their system identity. And in natural living systems these two models run simultaneously so both can change to come into congruence. Evolution may be rapid or slow, seeking this balance or having found it.

Now the case of an engineered system. The last thing the design engineer wants is something that modifies its existence and behavior. The engineer and market decides its existence/design and modifies that according to market success, but we don't what the engineered system to do it by itself. If your car changes into a toaster while you are driving, you would be very disappointed. So a goal of engineering something is to separate its operation from its design. But of course even with a completely physical machine we can't do that completely; all systems will re-design themselves to some degree, which is why there is always a "mean time before failure". But if we include the designer and factory, in fact it does behave like an evolving species. Many companies even take individual failed products back to analyze the failure and correct it - a kind of artificial selection.

In ecology, the boundary is very blurry now. It used to be preserved, but no longer with the idea of "niche defining phenotypes" (Odling-Smee), which is now recognized. In other words, many organisms (if not all to some extent) engineer their environment. A bird or gorilla nest may differ in sophistication from the Stanley hotel, but with regard to purpose they are more alike than different. There are also organisms that 'borrow' artifacts (e.g., hermit crabs). And there are humans who live very close to the land. So, where's the boundary?

From what I can tell reading Rosen, he was very keenly aware of the epistemic problem - that we only know our own experience and there is no direct knowledge of an outside world (also Descartes point in his discourse on method). The modeling relation expresses that idea. What we directly experience are the percepts in our own experiential apparatus. However, I don't think he meant to equivocate on the existence of an outside world. The point in understanding science is what we can know (epistemology), but

[Janet Singer]

John – Where are you that Michael's 10:55 pm email got time-stamped 11:25 am?

Janet

[Jack Ring]

JANET, 

Where was Michael when his email was stamped Aug. 24? 

[Duane Hybertson]

Hi John - Thanks for your thoughts, both the long version and short version. I have a couple of comments.

1. Patterns: You discussed systems having internal models, and mathematical descriptions of generic systems. I suggest that the concept of patterns--general models that characterize classes of systems--is a useful construct. Patterns can be observed in all types of systems, both human engineered ("artificial") and otherwise ("natural"). The conduct of both science and engineering depends heavily on patterns.

2. Views: You discussed incompatible models and dualities such as wave - particle. One practical (if not theoretical) way of dealing with these is the concept of views, which has been used to some extent in engineering but I think has much untapped potential as a tool for dealing with complexity.

3. Self-adaptation: You distinguished engineered systems by saying we do not want them to modify their own existence, behavior, or design, and that a goal of engineering is to separate operation from design. Those distinctions may hold traditionally, but they are becoming more blurred. I think systems engineering is moving toward a position of addressing more complex and "wicked" problems in society, and that requires producing complex solutions or systems--including a combination of people and machines--that can adapt to continually changing environments. [Software is an important element of adaptation on the machine side; we have had self-adapting software for over half a century.] I call this "collective actualization" - the combination of what the engineer anticipates and designs, and the ability of the solution system itself to adapt - instead of simply calling it "engineering". I also believe that to reach this more ambitious goal of systems engineering requires a greater understanding and incorporation of systems science. Which is why I am part of this SSWG.

Thanks,

Duane Hybertson 

--

[Duane Hybertson]

I should have mentioned that the importance of patterns in science includes systems science. SS patterns include the broader more general "isomorphies" that apply across domains and disciplines. Len Troncale's work on systems process theory exemplies this.

Duane

[David Rousseau]

James,

 

If I understand Rosen correctly he thought of “natural systems” as the systemic models we create when studying nature, so on his view there are “abstract natural systems” such as you refer to.  I disagree with him on that (see below), but I acknowledge that this is a bit of a philosophical minefield.  David Lewis in 1986 outlined five different ways of making the distinction between abstract and concrete particulars, and the debate is still ongoing amongst what Len calls “philosophers and other devils” :D

 

Here’s my take.  Many ‘abstracta’ can be considered to be “real” (exist objectively, hence part of Nature), e.g. numbers, functions, sets, propositions, relations, but we cannot ascribe causal powers to them without conflating them with the ‘concreta’ e.g. space-time, material substances and events.  I think that talking about “laws governing the behaviour of planets” or “feedback loops ensuring stability” are pragmatic ways of speaking that should not be interpreted as implying that laws and models have causal powers in addition to the causal powers of the substances whose behaviours we characterise using such laws and models. 

 

In line with the Rosennean modelling relationship (which I do like), I think that what we call “laws of nature” are abstractions we create in order to codify the “constant conjunctions” we observe in our experiences of natural phenomena.  Likewise, systemic mechanisms are abstractions we create in order to try and understand/predict the behaviour of complex material structures. 

 

However, as a moderate scientific realist I think that the successes of our best theories, and the congruence between these theories, give us grounds for believing in the real existence of the ontological entities our best theories postulate – so we can come to have knowledge of the world as it is “in itself”, Kant and Hume notwithstanding.  But the appearance of abstracted “laws of nature” and systemic mechanisms in our best theories only entail that there is a concrete world of substances with inherent causal powers, and that this concrete world really is organised in an inherently systemic way.   It does not mean that our abstracted laws and models are part of the objective reality we are trying to study and understand, or that they can interact with the concrete world or with each other (for all that it is sometimes convenient to speak that way).  But it does mean that if we do our work well, the logical coherence of our models will match the causal integrity of the concrete world.  Our rational balancing act across this Rosennean bridge is also a systemic process.  This emphasises the importance of systems literacy:  knowledge of the nature of systems helps us not only to build good theories and to understand the nature of world but also to assess the quality of the match between our theories and the real world.

[Jack Ring]

We should be cautious about isomorphies. An isomorphy exists based in a limited number of attributes although not explicitly listed.

We should consider them as hypotheses until we do science by vetting them for dynamic and integrity limits. 

Everything with wings is not an airplane. 

[Hillary Sillitto]

David

This is indeed a minefield. The good news is that for me at least, the discussion over the past few days has made up for months - nay, years - of frustration, when this discussion forum seemed not to be getting anywhere. Now it seems we have taken a leap forward, (no doubt enabled by the years of less satisfactory but probably very important precursor discussion), triggered by your manifesto and greatly aided by some of John's carefully phrased comments.

I think your analysis above is close to the mark. I would add a further insight from work done by David Blockley at Bristol a number of years ago. He maintained that "a system is a process". I confess I found his argument hard to take on board at the time. However, having observed this thread of discussion and also the discussion thread I started in Systems Thinking World a few weeks ago, it starts to make sense. The short version:

Rosen refers to a natural system as a mental construct because, like the scientific models and theories you refer to, Rosennian "natural systems" are descriptions of our best current understanding of the elements and interactions that we believe account for the observable behaviours of the natural world. The observable behaviours result in effects that we can perceive - Rosen's "percepts". Now taking this in reverse: The "percepts" - what we can perceive - are the effects produced by natural processes. The effects are observable, and our observations allow us to infer the existence of the natural ("real-world") processes that produce the effects we can observe. So the evidence for the existence of systems in the real world is not the observation of "things" that we would label as "systems", but the observation of effects produced by processes that cannot be ascribed to single real-world "things" acting alone. So what we can perceive in nature is not "systems" but "systemic behaviour". So our primary tool for establishing correspondence between conceptual and "real" systems is not "things" but "effects". "Effects" that cannot be attributed to individual "things" are evidence for the existence and relevance of "natural systems".

Practically every observation we can make about nature is actually to do with a process rather than a thing. We only see "things" if they reflect or emit light (a "process"). We often don't see interactions between "things" (animals eating plants and other animals being an obvious exception...), we can usually only infer interactions between "things" indirectly, by the effects they result in, and by what happens differently if we interfere with the process.

So a Rosennian "natural system" is that set of things and their interactions which we hypothesise is both necessary and sufficient to account for the effects we observe. So a natural system of interest is defined - essentially reverse engineered - to explain a natural phenomenon, the "effects" produced by "process" performed by the "system of interest". The system of interest is defined by the effect of interest. By contrast, a man made system is forward engineered to produced a desired effect (and, we hope, no other undesired ones, but that's the difference between good and bad systems engineering). Again, the system of interest is defined by the effect of interest. 

So this - very crude - set of ideas offers a way to connect the Rosennian world to "objective reality". It, or a better but similar formulation, may already be obvious to others, but setting it out in a way that makes sense to a wide range of stakeholders would take us a big step forward.

In terms of a manifesto and "what good looks like" as we move forwards: at the moment the language is pretty impenetrable. It needs to get closer to stakeholder language. External stakeholders who should have an interest in this discussion will not understand the depth of meaning and history behind such phrases as "the epistemic cut" or "the modelling relationship"; and they will be totally bamboozled if and when they discover that a "natural system" is not a natural system at all but a mental construct.

By the way, it's necessary for the "system literacy" initiative to take account of at least three statements that are often taken to be fundamental truths, and often used as blockers to useful discussion, which this discussion thread shows need to be taken much more carefully than that:

1. "systems are purely mental constructs" versus "systems exist in the real world" - negotiating a route past this roadblock seems to be within reach as a result of this discussion.

2. "systems have purpose" - hasn't been such an issue recently, but clearly only applies to man made systems unless you take a very wide view of "purpose"

3. "systems have boundaries" - certainly true of engineered systems and of system models, but I've often found it useful or necessary to defer discussion of boundaries in systems until a systemic understanding has been gained of the situation. Premature and too-narrow definition of boundaries seems to me to have been the primary failure mode of many unsuccessful attempts to create or improve systems.

I will look for my Lighthouse material later today.

Hillary

[Jack Ring]

David, 

I suggest that James’ characterization of Rosen’s idea is not correct, specifically, “…Rosen's formal system per this construct is basically our thoughts about a system that exists in the world (or one that might or could exist).…" stated Rosen incorrectly. 

IMO formal means ‘according to vetted rules of logic, mathematics and semiotics’ which means a formal system cannot exist in our thoughts (the human mind cannot discern reality from illusion), it can exist only when expressed in a media which can be shared by others thereby enabling the vetting of the hypotheses.

As proposed in the SysSciWG ontology project a couple of years ago ‘system' can exist a) in tacit thought, b) as expressed in descriptive or prescriptive models in various languages (natural, graphical, mathematical, choreographical, music notation, playwright conventions, Java, etc.) including various degrees of mis-communication, and c) in media having mass, length and time attributes.

The human hubris to call an ant colony or the configuration of planet dynamics a system is an exercise in mapping, not system science.

This view is not popular with those who seek to devise a philo rubric for sophistry but it sure helps keep buildings from falling down.

Make sense.

[Janet Singer]

Hillary

This is a very helpful exposition in broad language of the basic resolution of apparent conflicts that is available if we are careful with language and can integrate the unifying process-oriented pragmatic insights from 19th c science and philosophy of science that were lost in 20th c over-specialization. When I argued for a process-oriented foundational view at Linz in 2012, James in particular had a strong allergic reaction given the special (and negative) associations with 'process' in SE. Len has encountered the same problem with his Systems Process Theory label, arguing unsuccessfully that process orientation is the default view in science (and likely the working approach of most engineers as well).

The challenge is integrating the insights that have long been available and, as you note, communicating them in accessible language relevant and most useful to the various audiences and contexts.

(Effects-based thinking is relevant as well.)

Janet

[Janet Singer]

Jack

I'm not sure if Rosen said explicitly that a formal system cannot exist in our thoughts, but the idea that a formal system needs to be 'vettable' or socializable is certainly something that Peirce emphasized.

[Jack Ring]

it may be useful to discern the difference between a) process and b) output, outcome and effect, the second set being results of process.

Until we see system as both 'particle and wave’ what it IS and what it DOES we will wander in the sophists wilderness another 40 years.

[Jack Ring]

I’m > 2% sure that he meant a formal system exists primarily in our thoughts and sometimes in our handiwork (but that version usually includes logic, arithmetic or semiotic errors).

Judith?

[Lenard Troncale]

All,

I am moved to respond to Hillary's recent statement, especially the nicely concise summary 3 statements at the end which have frustrated me, and us, so much in the past. (1) "systems are purely mental constructs" versus "systems exist in the real world" (2) systems have purpose, and (3) systems have boundaries.

In my upcoming book, I have a theme described in the "forward" with "theme boxes" peppered thruout the text. One key one, relevant to this discussion, is named "BOTH-AND not EITHER-OR" with nuances in it on duality, redundancy, and the unrecognized primacy of paradox in the natural world. These address exactly Hilliary's concise summary of the existing positions that cause endless debate.

I think the point is that for all (3) above, there is endless argument because there are such good arguments for both positions. Systems exist in the REAL world AND they are also mental constructs. BOTH-AND. As I have argued endlessly before, scientists vigorously oppose the use of the word "purpose" in the former, but if one substitutes the word "function," the dissonance disappears if people let it. For both sides of each, there is a special audience that is devoted to just one of the sides. Function subsumes purpose or at least better descrlbes it for the real systems. But, as Ring endlessly says, "what it does" is of great importance. So, again BOTH-AND. As I hope to explain in the book.

And as for boundaries, the basal, most accepted definition of system alludes to all parts being connected or acting on each other. IF everything IS connected, then humans viewing reality or their own produced systems can easily describe different boundaries for different intents or for capturing different elements for different functions/purposes, and so boundaries are fluid. SO again BOTH-AND. (Except in those rare cases where extensive experimentation has revealed the boundaries that nature uses to make the system fulfill its purpose ((WHOOPS! function)), and even those are conditional to later experimentation, and to know those explicit boundaries you have to be an EXPERT on those extensive experiments; many of our commentators are not and are just imposing their beliefs).

Humans don't like BOTH-AND or paradox. Also explalned in the book. SO the debate goes on.

Amen?   Len

see John Kinemann, I can rarely be brief too. ((Notice this sentence can be taken two ways too, BOTH-AND))

[Jack Ring]

Len, 

Do you distinguish Both-And from Having it Both Ways?

Perhaps Function is a Both-And because there is the function a system performs with respect to stimuli from its context (not system) vs. the functions performed by ‘components’ in the system.

Of course this components confusion can be avoided if one stops talking about components of a system because these are simply incorporated systems thus function with respect to stimuli from elsewhere in the system, not from the context.

Likewise most configurations are capable of responding to more than one stimulus so "different boundaries for different intents” simply says that a configuration is capable of manifesting more than one system.

Perhaps Rosen was encouraging us to think about types of systems as well as classes of systems. Philosophers, bio-logists and engineers tend to be mesmerized by the classes.

Jack

[Janet Singer]

Len,

To the extent that 'scientists vigorously oppose the use of the word "purpose"' it is because they are relying on cultural practices (forbidding taboo concepts and terms) as a means of quality control rather than trusting scientific method. 'Function' should allow the dissonance to disappear if people will let it, as you said.

Quality is better assured by keeping in mind

1) The pragmatic scientific worldview, as recently described by Hillary:

The "percepts" - what we can perceive - are the effects produced by natural processes. The effects are observable, and our observations allow us to infer the existence of the natural ("real-world") processes that produce the effects we can observe. So the evidence for the existence of systems in the real world is not the observation of "things"  that we would label as "systems", but the observation of effects produced by processes that cannot be ascribed to single real-world "things" acting alone. So what we can perceive in nature is not "systems" but "systemic behaviour".

2) Fallibilism, or recognition of need for ongoing vetting of hypotheses, and

3) Means for carrying out that vetting through innovation, experiment and public debate. (As Peirce pointed out, science is the means for clarifying beliefs that provides for ongoing obsolescence of its own accepted 'truths'.)

Carefully following (1), above, we don't observe real 'boundaries' but patterns of effects that lead us to hypothesize that some 'thing' or structure (slow process, as you have noted) is functioning as a boundary. So the notion of function is key for clarity in science as well as in engineering.

Janet

[Hillary Sillitto]

Lighthouse draft material is consolidated here: https://www.dropbox.com/s/5sbdozu5qx9x218/System%20lessons%20from%20the%20Scottish%20Lighthouses%2C%2018th%20to%2021st%20centuries.pdf?dl=0 - NB please feel free to read and comment, but not to distribute the pdf document outside this forum.

[Hillary Sillitto]

I'd like to make a further observation on the Rosen notions of "natural" and :formal" systems, in the context of the model that Kenneth Lloyd has proposed in his draft book on Category Theory.

Ken's model has four "domains" - real world, mental, conceptual, and the "platonic mathematical world". The first three are closely related to Popper's three worlds and the fourth to Penrose's - I think the full name is - Platonic world of mathematical forms.

It seems fairly clear to me that Rosen's "natural systems" belong in Ken Lloyd's "conceptual domain", corresponding to Popper's third world of shared knowledge. (The notion of a particular natural system will originate in the mental world as an individual's mental model; once shared and described in document or other form, a "natural system" would move into World Three, the world of shared human knowledge. By Rosen's definition it does not belong in the "real world".)

Rosen's "Formal system" has to be self consistent and in some sense mathematically defined and constrained, so it belongs in the Platonic Mathematical World (I don;t feel comfortable with the terminology but am sticking to Ken's language as best I can to avoid spreading further confusion by possible erroneous interpretation.)

Rosen's "Material Systems" exist in the real world, Popper's World 1. As discussed above, their systemicity may not be directly observable, but inferred through the effects they produce.  

Even if this doesn't explicitly match Rosen's definitions, I think there is value in using Ken's formalism to help clarify and advance our thinking on this. 

Hillary

[Lenard Troncale]

Jack,

I have no problem with phrases like "Having it Both Ways." Nature does this all the time. In human discourse though this phrase is often derogatory.

My response to your insight about components would have been exactly what you say in your second sentence of the second paragraph.

I might substitute "set of components" rather than "configurations" in your third paragraph, but essentially this has the same meaning.

It is interesting that I would use "classes" and "types" as similar words because most humans tend to classify types, thus leading to classes.

Will we ever get past our slavery to words? Yes, Hillary and Rosen and modelers would say that using math and simulations will. But I would say a lot of words have to be disciplined by the results of lots of experiments before the words give us the chance to model and simulate, and sometimes word insights actually lead to math advances.

Len

[Lenard Troncale]

Janet, 

I do agree with all you say here and it sounds like at least we two are in substantial agreement. I would ask Hillary to include "things" as well as "processes" but as author of Systems Processes Theory" I can hardly disagree with the primacy of processes. I think it is okay to see mitochondria as bounded entities, or cells, or organs, or galaxies, or a solar system. Each has "thingness" and a relatively well defined boundary for the processes that count.

I have also noticed that humans first recognize "thingness" in a primitive science. Objects are detected and named in chemistry, geology, biology long before they are subjected to investigation. Objects are easiest for us to see; it takes much longer for us to investigate and understand dynamics. But both are important as developments in a new science.

Len

[Janet Singer]

Len,

I think it's fine to use concepts and labels like 'thing', 'system', 'component' as long as their meanings are clear from context. But any time there is a need for exceptional clarity – such as when trying to  communicate across disciplines, cultures, technologies – grounding in observed regular patterns of effects is the best hope for disambiguation.

Of course 'observation' is a perspective-dependent and context-sensitive process so this hasn't eliminated the challenges. It has just relocated them from pointless arguments about ontic status to questions we have some hope of answering as science progresses.

Janet

[Jack Ring]

Hillary, Janet, Len, et al,

Why does Hillary find it necessary to specify that effects come from process? As the bumper sticker says, “Effects Happen”

And, if process is somehow necessary then is it a) the process that manifested the effect or b) the process by which the observer decided which process was really “the process that manifested the effect” then vetted his/her choice?

However, if you really need process then please go back to the point ignored a while back and describe the ‘catalyst' process.

I encourage the use of system to mean “that which produces acceptable responses to authorized stimuli” then using another label for the ‘that’ when it is not responding to stimuli. I suggest the use of configuration. I think this distinction gets at the Rosen and Pierce notions of mental vs. physical but also vs. temporal.

Is process a thing or does it somehow predict a transform? Is a player piano role or Jacquard’s string of punched cards for knitting looms a process or just a program? If program than what means process, something static or moving? If moving then what shall you call it when not moving?

Yes, humans first recognize thingness (Cabrera’s Distinctions?) but as early as the beginning of the third trimester, humans recognize not only different tones but also timbre, chords and melody (Cabrera’s S, R?) and respond to selected examples (Cabrera’s P?).

Yes, Ken’s four factors, real world, mental, conceptual, and the "platonic mathematical world” merit attention. It may be that all the “formal” logic, arithmetic, algebra, calculus, semantics, music, semiotics, etc. fit in the "platonic mathematical world.” 

And we now know that emotion is a key factor. During designing, System Thinking is rarely sufficient without System Feeling. Then System Doing tells us how naive we were.

Onward,

On Aug 24, 2015, at 2:21 PM, Lenard Troncale <lrtro...@cpp.edu> wrote:

[Jack Ring]

Just as "Science progresses one funeral at a time.” I think Systemics will progress one birth at a time. if you have not experienced the 4 to 10 year old kids encountering DSRP and other kinds of systems thinking, feeling and doing then I encourage you to do so. These kids are not sophists.

[John Kineman]

Janet,

I'm in South India

John

[John Kineman]

Hi Dwayn,

All great comments. Agreed.  I also use the term "pattern" but was undecided if it belonged with exemplar or design. I think it belongs more with exemplar because that is a previously defined physical structure that has formative effect in context. So, the landscape in Alaska provides an exemplar for patterning the migration of Elk. But it does so by entering first into contexts relevant to the Elk, like the implications of the landscape for energy expenditure from topography, or movement resistance due to substrate, etc. The physical pattern exists first, then its implications for these formative contexts, which then determine the Elk dynamics of migration. So I think pattern is best associated with final cause (in a naturalistic sense). It works out nicely that way in the quadrant diagram too, because then you have Pattern and Process opposite each other, so assuming the intervening causes, it becomes one approach to understanding the whole cycle. And that is what landscape ecology is based on -- pattern and process (Forman and Godron 1981).

'Views' are extremely relevant I think, but again I'm not sure where they fit in the schema and if the concept is clear. What would be the relationship to "world view" on the one hand, and "frame of reference" on the other? A frame of reference (arguably an observer's 'view' point, as in relativity) exists as multiple possible relativistic frames within the single 'world view' of relativity. Like so many good concepts, we might be using it for many things that need to be distinguished.

Good point about engineering adaptive or intelligent systems - I was thinking about design of machines, like a car or a toaster. Would you say that the machine concept begins to blur as we design more adaptive and intelligent systems? I wonder if there might be something important about how we do the adaptive or intelligent design. Robert Rosen, and certainly Judith, make a strong distinction between 'simulation' and 'model', and there has been a big question around von Neuman's concepts of intelligence (Rosen thought they were on the side of simulating intelligence or complexity rather than modeling their true nature). The issue is if we are making more intelligent systems or better simulations of intelligence, in which case it would still be strictly a machine. I think this distinction is a very difficult and controversial one right now. I think of my 'smart' TV. It simulates intelligence, but there is no reasoning with it! It seems like the worst of both worlds. But certainly in systems design where people or other living beings are involved (true complexity and intelligence) the autopoietic nature of the engineered system is not a simulation, but real complexity. So I accept your point.

Cheers,

John

[John Kineman]

Duane,

Again, agreed.

[Duane Hybertson]

John - Thanks for your response. Regarding views: I see view as a very general and flexible construct that can include world view and frame of reference as types or examples. In fact, I suggest consideration of the possibility that view is as fundamental to systems as set theory is to mathematics.

Duane

[Jack Ring]

 Regarding engineered systems, I suggest that we are already deploying ones that modify their own existence, behavior, and design, c.f., Sysfems of the Third Kind, INCOSE INSIGHT, July 2014

[John Kineman]

Hillary, Wow - great synthesis which I won't soil it with comments.

On the three action items:

1. The holon view says both are true and in relation to each other, if we broaden the concept of 'mental' to 'contextual'; i.e., contextual constructs that are realized in the material world.

2. Yes, need to broaden the meaning of 'purpose' - it was the anthropomorphic metaphor for Aristotle's final cause, and left as a hierarchical penultimate 'final' it was taken to mean Immanent or Divine cause. The naturalistic equivalent is fitness to a prior exemplar. With that we can get rid of the hierarchy by making it a closed cycles of causation, because 'prior exemplar' comes from the lowest level of the previous hierarchy, material states or measures. In the hierarchical view we dispensed with final cause as best as possible, but it was not possible to dispense with it completely (because, I believe, it is not a hierarchy). Thus it was reduced to only one possible purpose - survival. In fact probably no organism but thinking humans considers survival as such - they are interested in food, sex, shelter, landscapes, and life strategies -- a myriad of purposes encoded into the internal models.

3. Need to distinguish clearly between physical boundaries (skin, shapes, surfaces, etc. that co-locate the system), and causal boundaries (final, formal, efficient, material boundaries whose openness or closedness define life, complexity, impredicativity, intelligence, system identity, etc.) that define the organization of a system, along with their relations and permeability. For example, two different religions are separated by a formal boundary (different belief systems), but may co-locate anywhere in the world.

Cheers,

John

[John Kineman]

According to RR (Judith can confirm), the modeling relation could be applied even within mathematics itself, to the study of number theory, for example. It is an entirely generic template. Similarly it can apply entirely in the natural world, between two systems in nature, although that can be known to us only by inference because of the epistemic problem.

The best way to understand this flexibility, that I've found, is to think of "Natural System" as meaning the objectified aspect of a system and also "Formal System" as the formalized aspect of a system. Indeed, he said it is necessary to consider both 'real' (why I choose to say 'quasi' real, rather than 'moderately' real - there is nothing moderate about this view). The relation is 'impredicative' in nature (we think), meaning that all systems are fundamentally complex. While the modeling relation shows just one idealized relation, actually it is an infinite holarchy. RR said this in the statement "there is no largest system" that can serve as a formal representation of any natural system, unless we say that things and mechanisms are fundamentally real in and of themselves, which of course, RR argued against. The object and the mechanism are appearances in a world that is reduced to apparently consistent natural laws, but we have seen the holes in that view sufficiently now to know it is a reduction of the complex, not an absolute 'concrete' reality.

Jack - I would encourage a broader view than implied in your call to "make sense". Sense is like beauty, in the eye of the beholder, hopefully many of them together so it is a shared sense, but still problematic when considering new ideas. Consider for example, Richard Conn Henry's 2005 paper disclosing the tacit agreement among physicists, stated thus (article attached):

"The only reality is mind and observations, but observations are not of things. To see the Universe as it really is, we must abandon our tendency to conceptualize observations as things."

Now I don't entirely agree with Henry and I'm talking to him about that. I think we can't call the mental 'real' either. Each side, the contextual (mental) and realized (material) are real in relation to the other. We got used to the material reality because we grew up in a very stable and general context, but when we step out of that general world context, say into each person's mental universe, or quantum reality, we find that its a relation between contextual/mental universes and their exemplified and formally bounded objects; i.e., modeling relations.  So, rather than modeling relations being such a radical and abstract idea, if you consider where science actually is today, modeling relations are the compromise.

[Jack Ring]

John, 

Apology. I failed to include the ‘?’ 

I was not calling for agreement. 'Make sense?' was intended to trigger a conclusion on the reader’s part when considering new ideas.

[John Kineman]

Len, yes, "Both-And" is practically my mantra. I wholeheartedly agree. And Paradox is a very under-appreciated concept. I personally think 'reasoning by paradox' is the way to real progress in science. I hope you give it prominent treatment in the book. There was a book I had long ago by Rorlich that, while a bit too fixed in realism, was nevertheless quite instrumental in helping me with this concept. It is: Rohrlich, F. (1989) From paradox to reality: our basic concepts of the physical world, Cambridge University Press, Cambridge; New York.

John

[Hillary Sillitto]

Jack

Your question re process. I view process as something that happens. By deferring association between process and object we defer the argument about identifying the system. By associating process in the material world with observable effect, we make the link between the material world and our observation and interpretation of it. I view a system as a set of interacting parts exhibiting emergent properties, properties not attributable to any of the individual parts acting alone. So if the observed properties are not attributable to any one object, they must be due to interactions between multiple objects - hence, due to a system. So we can deduce that the 'system' needs to exist to explain the observation, even if we can't as yet identify it, or the configuration needed to instanciate it.

Hillary

[John Kineman]

Hillary,

Rosen's modeling relation is between subject and object, nothing more esoteric than that. The first example of it was science itself, so clearly 'Natural System' can be anything that science studies. 'Formal System' is the domain of models, which exist as or in mathematical 'formalisms' (taking the broadest definition of mathematics). Applied generally, it can be very inclusive of any 'view' as Duane suggested using that term. Those two domains "Natural System" and "Formal System" would then cover all there is, except for a very interesting fact of the modeling relation -- the encodings are not in either box. Where are information processes (encoding and decoding)? That is actually where the knowns are. We can know the measurements (observables) and our percepts (decodings). The Natural and Formal systems (which are really one system) constitute the ontological aspects of the relation; observables and percepts constitute the epistemological aspects. A system, in this view, is both natural and formal -- Both/And. Inference is as real as causality (Aristotle called both "aitia', which I think is the root of our english word ending -ation, which means some kind of effect).

We should not be thrown by his defining 'natural' in terms of observables. That simply recognizes the fact that observables are all we can know of an event in nature. Thus, any definition of a natural system has to be in terms of its observables "for us" (because that's all we can access of a natural system). It is strictly a technical epistemic matter that we have to remain agnostic about the 'concrete' existence of something and speak only through observables and percepts. This kind of logical precision allowed him to say things later about the limits of science and where it should go.

So natural systems can be trees, cars, people, ecosystems, climate systems, governments, cells, organisms, phone books, religions, policies, economies, atoms, particles, waves, folding processes, etc. They can even apply to mathematical objects themselves. It is anything one might make a model of (or more broadly, create knowledge of). It is a representation of what is experienced.

But there is a catch. Its a bit of a digression from the main topic here, but RR felt strongly that the model should not merely represent appearances, but actually replicate the entailments in nature, i..e, the way the system works in nature, using formal logic. He saw a sharp distinction between something that 'looks and quacks' like a duck, and something that is causally entailed as a duck. This is the only case I've found so far that I thought might not be logically iron-clad. I'm personally not sure how one can really make this distinction, except as a matter of intent of the investigator. In other words, if Ptolemy really believed nature was entailed as circular motions, then he was making a model. Later we find out there is a better idea of how it is entailed, so we shift to the Newtonian view. Now Ptolemy's system is a simulation, and Newton's is a model. Then we find that Newton's idea is also limited in some way, so now his model turns out to only be only a simulation. It seems to me this is the realism problem, but I agree with RR's distinction as far as goal and intent. Maybe its also possible to say that Ptolemy was "wrong" because in fact there are no circular motions in nature. Even accepting approximation and error, the entailments are not constrained at any level so that objects move in circles. Whereas Newtonian forces seem to survive better, but now it seems to be getting into Jack's solipsism. So, maybe, I'm not sure, it is possible to state a method for verifying the entailment hypothesis. In any case I don't suggest discussing this, because it may have no useful end - the debate between real and artificial intelligence is a case in point.

John

[John Kineman]

Thanks Hillary, that's useful.

My understanding of Process Philosophy/Theory (Whitehead) is that it is a radical shift much like relational theory from the view that objects are real and their properties are secondarily real; to the view that a highly generalized 'process', which is systemic in this "emergent" way that Hillary describes, and can involve mental as much as material aspects. It is responsible for generating the sense of an object that we know through its properties (Rosen's observables). So maybe that's a different sense of process than is common in current science, where it is usually restricted to mean dynamical process.

JohnK

[David Rousseau]

Hillary,

 

Thank you very much for this very clear exposition, and your general guidance about ‘pressure points’ for this discussion. 

 

Your clarification of Rosen’s categories, together with your later post linking the Rosenian categories to Ken Lloyd’s, provide a very helpful map for navigating these concepts.  I think the conceptual categories are very useful but the naming less so, and this speaks directly to your point that we have to get closer to stakeholder language.  Of course we have to figure out clear concepts first, before we can work out accessible ways of speaking about them, but we should start at soon as possible.  One way is to try and stay as close as possible to widely accepted meanings, and to try and avoid developing new technical terms, especially by embracing unusual meanings for established terms.  In this sense I find Rosen’s terms not very helpful, and I think we should be prepared to revise them.  If we want to allow for the possibilities that there are systems in nature and systems in our mental world it seems odd to call the systems in our mental world “natural systems” and then have to cast around for a term to refer to the systems in nature.  Calling them “material systems” provides a theoretical solution but the price is very high, since it is out of alignment with the vast body of philosophical literature in which the material world is synonymous with the natural world.  It would be much more intuitive to call the systems in nature “natural systems” and the systems in our minds “conceptual systems”.  Rosen is sometimes misinterpreted in this intuitive way, see e.g. the description in John Barrow’s 1998 book “Impossibility: the Limits of Science and the Science of Limits”, p. 194, where he draws the modelling relation as spanning between (objective) phenomena in nature and mathematised theories about them. 

 

Your discussion about processes is very opportune.  I concur with the sentiment expressed in the claim “a system is a process” but David Blockley’s phrasing is infelicitous here.  It would be better to say something like “a system is inherently dynamic”, or “all concrete properties are conditioned by systemic processes”.  That way we include a focus on change in our analysis, we do not imply that we are providing a whole or a fragment of a definition, and we don’t conflate the useful categories “system” and “process”.  As an aside, when Ervin Laszlo wrote “Introduction to Systems Philosophy” he claimed that SysPhil is the next step in a philosophical progression that has Whiteheadean Process Philosophy as its predecessor (p.12).  I think he was right in this, although this idea is yet to be developed properly.

 

With warm regards,

[Janet Singer]

All,

Whitehead is the best known Western proponent of process philosophy/theory (or 'philosophy of organism', as he termed it).  Whitehead was a mathematician prior to his work in philosophy at Harvard, and the motivation for his philosophy of organism was to develop an integrative foundational framework that was consistent with both modern science and our common sense experience (or 'radical empiricism' in the words of William James, Whitehead's colleague at Harvard).

Whitehead's work had and continues to have a major influence on the development of systems science beyond his influence on Ervin Laszlo that David noted. To give just two examples, Jim Miller was a student of Whitehead's at Harvard, and it was Whitehead who suggested Miller pursue his unifying Living Systems Theory; and Warren McCullough, student of Rashevsky's (as was Rosen), pioneer in neural net theory and one of the founders of cybernetics said that the inspirations for that work could all be found in Whitehead's Process and Reality. 

From the Stanford Encyclopedia of Philosophy:

The perhaps most powerful argument for process philosophy is its wide descriptive or explanatory scope. If we admit that the basic entities of our world are processes, we can generate better philosophical descriptions of all the kinds of entities and relationships we are committed to when we reason about our world in common sense and in science: from quantum entanglement to consciousness, from computation to feelings, from things to institutions, from organisms to societies, from traffic jams to climate change, from spacetime to beauty. Moreover, results in cognitive science, some philosophers have claimed, show that we need a process metaphysics in order to develop a naturalist theory of the mind and of normativity. These arguments form the background for the processist criticism of the focus on substance in Western philosophy.

The bias towards substances seems to be rooted partly in the cognitive dispositions of speakers of Indo-European languages, and partly in theoretical habituation, as the traditional prioritization of static entities (substances, objects, states of affairs, static structures) at the beginning of Western metaphysics built on itself. In contrast, process philosophy shows fewer affinities to any particular language group and can allude to a rich tradition of reflection in many of the great schools of Eastern thought. Thus contemporary process philosophy not only holds out the promise of an integrated metaphysics that can join our common sense and scientific images of the world. It is also of interest as a platform upon which to build an intercultural philosophy and to facilitate interdisciplinary research on global knowledge representation by means of an ontological framework that is no longer parochially Western. http://plato.stanford.edu/entries/process-philosophy/

Janet

[Lenard Troncale]

All,

Janet's exposition causes two responses in me. One perhaps supercilious and one merely a sincere anecdotal sentiment because Janet, in her last msg, outlines a connection between Miller and Whitehead.

First, I feel myself fortunate now that I termed my lifework "Systems Processes Theory" although it was rather independent of knowledge of Whitehead. I do not read in philosophy and am ignorant of the details Janet so nicely illuminates here. My emphasis on process comes directly from the emphasis on process in the natural sciences, and science's relative success in elucidating processes thru experimentation.

Second, one of my favorite anecdotes of many conversations with Jim Miller is the one after the following. (it may not be remembered that Jim hired me for a summer when he became President of University of Louisville to help him found his Systems Science Institute there. He told me at our first meeting in his grand Presidential office, with a copy of Rodin's The Thinker nearby ((first casting)), "your major obstacle is my Provost, Dr. John Dillon. He is a physicist and thinks we may be wasting funds on a new Institute focused on living systems." I created an Institute Fellowship for him at U of L (imitating mine founded at Cal Poly Pomona in 1972) and the Fellowship started meeting that summer.  Well, John was also a hard Irish drinker, pipe smoker, and loved to party and to argue. He loved the Fellowship idea and became one of our most ardent debaters and supporters that summer. Later he became the 2nd Director of the Institute, and eventually even President of the ISSS ! His current Wikipedia entry says he was the Founder of the Systems Science Institute. Not true. He was first its opponent and only later its savior. Ironic. But that is not the anecdote).

James once told me that when he was very young, his father was President of a University and he always had that ambition in mind for himself. It was now fulfilled as he became President of University of Louisville. Because his father was a university president, their family house had many famous visitors. One of these was Alfred North Whitehead when he was still a small child. Whitehead quipped while holding Miller on his knee, "when I was young Charles Darwin bounced me on his knee, and James would add that Whitehead bounced Jim Miller on his knee." So he felt he had continuity all the way back to Darwin himself. Pretty impressive. Some were turned off and criticized by Miller that he would tell stories like this that they felt aggrandized himself, but I always found them simply and sincerely delivered with a sense of awe that created the same reaction in myself. It was a time of giants living their natural lives in networks.

[This anecdote will be in one of my "theme boxes" in the upcoming book, but you heard it first here on SSWG Google].

Len

[Janet Singer]

Len,

That is indeed an awe-inspiring story.

In honor of 'that time of giants living their natural lives in networks' (as you eloquently put it) I am sharing a sketch of mine from 2012 showing some of the connection I had noted.

On Aug 25, 2015, at 9:27 PM, Lenard Troncale <lrtro...@cpp.edu> wrote:

All,

Janet

[joseph simpson]

Nice...

--

The SysSciWG wiki is at https://sites.google.com/site/syssciwg/.
 
Contributions to the discussion are licensed by authors under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.
---
You received this message because you are subscribed to the Google Groups "Sys Sci Discussion List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to syssciwg+u...@googlegroups.com.
Visit this group at http://groups.google.com/group/syssciwg.
For more options, visit https://groups.google.com/d/optout.

--

Joe Simpson

“Reasonable people adapt themselves to the world. 

Unreasonable people attempt to adapt the world to themselves. 

All progress, therefore, depends on unreasonable people.”

George Bernard Shaw

[Lenard Troncale]

Janet,

Your chart summarizes so much. All it needs is a short paragraph of detail for each arrow explaining the relation concisely to be multiply useful for folks like myself who need the explanation. As a learning device. Also it would be wonderful if you could interact with folks like David Rousseau and/or Kent Palmer who study similar networks to come up with a consensus graphic. Thanks very much.

Len

On Aug 25, 2015, at 10:34 PM, Janet Singer wrote:

Len,

That is indeed an awe-inspiring story.

In honor of 'that time of giants living their natural lives in networks' (as you eloquently put it) I am sharing a sketch of mine from 2012 showing some of the connection I had noted.

<Pragmatic Roots Singer 12-0904.jpg>

[David Rousseau]

Janet, brilliant!  Thank you! J J

 

I like Len’s response too, but he’s being too timid.  Your diagram includes (if I counted correctly)  36 names, 81 arrows and 13 “traditions”.  A paragraph on each would add up to a book, but what a fantastic contribution to our field that would be!  Please, will you do it?? Forget consensus with me and Kent (or anyone), give us all a baseline asap!

 

David

 

From: syss...@googlegroups.com [mailto:syss...@googlegroups.com] On Behalf Of Lenard Troncale
Sent: 26 August 2015 16:24
To: syss...@googlegroups.com
Subject: Re: [SysSciWG] Definition or Types of System

 

Janet,

[Richard Emerson]

Janet

I agree with Len and David, short expositions of the each of the entities would be a great book for introducing  systems science, guiding the study of it and suggesting further research in it.   

Your diagram is also a great framework for developing questions to ask.  A thought provoker.  I think you could keep an army of PHD students busy for years, and not only busy but productive in both understanding the development of the field and pushing the boundary.  For example, what other systems science concepts could be included.

Brilliant

Dick 

[Lenard Troncale]

David, All:

I have rarely been characterized as timid, David. Thank you. Another advantage of this book "baseline" would be that it could and should be built upon and its existence promotes and enables building upon.

For example, Hierarchy is represented by many more than Allen. There are nice published summaries by Wilson, Wilby, Salthe, Simon, Pattee, Miller, Bonner, Eldredge, Odum, Mesarovic, Weiss and more that are at least as penetrating and detailed (book length) as Allen. I myself have 27 products on Hierarchies. Existence of the baseline allows stable addition of extensions that need to be known. Also there are many more than 13 "traditions." In any case, what exists already -- the at-a-glance summary graphic you have produced --  is very useful.

Len

[Jack Ring]

Hillary, Janet, Len,

Why does Hillary find it necessary to specify that effects come from process? As the bumper sticker says, “Effects Happen”

And, if process is somehow necessary then is it a) the process that manifested the effect or b) the process by which the observer decided which process was really “the process that manifested the effect” and vetted his/her choice?

However, if you really need process then please go back to the point ignored a while back and describe the ‘catalyst" process.

I encourage the use of system to mean “that which produces acceptable responses to authorized stimuli” then using another label for the ‘that’ when it is not responding to stimuli. I suggest the use of configuration.

Is process a thing or does it somehow predict a transform? Is a player piano role or Jaquard’s string of punched cards for knitting looms a process or just a program? If program than what means process, something static or moving? If moving then what shall you call it when not moving?

Yes, humans first recognize thingness (Cabrera’s Distinctions?) but as early as the beginning of the third trimester, humans recognize not only different tones but also timbre, chords and melody (Cabrera’s S, R) and respond to selected examples (Cabrera’s P).

[Jack Ring]

Where are John Warfield, INCOSE Pioneer and GSR Pres. and A. Wayne Wymore, INCOSE co-founder?

On Aug 26, 2015, at 9:59 AM, joseph simpson <jjs...@gmail.com> wrote:

Nice...

On Tue, Aug 25, 2015 at 10:34 PM, Janet Singer <janetm...@gmail.com> wrote:

Len,

That is indeed an awe-inspiring story.

In honor of 'that time of giants living their natural lives in networks' (as you eloquently put it) I am sharing a sketch of mine from 2012 showing some of the connection I had noted.

<Pragmatic Roots Singer 12-0904.jpg>

[Janet Singer]

Len,

So after the initial burst of enthusiasm we move on to pointing out the omissions?? You, of course, have been facing the same problem: when is an innovative overview ‘good enough’ to get out to the public, when the author can see it is so obviously incomplete?

I guess the lesson when dealing in uncharted and cross-cutting ‘hybrid territory’, as we are, is to use engineering rather than a scientific or scholarly standards, and an agile rather than waterfall approach. The iterative improvement needs to start somewhere (as David and Dick helpfully framed it).  

I’ve been saying similar things to you for years – time to take my own advice?

Janet

On Aug 26, 2015, at 9:51 AM, Lenard Troncale <lrtro...@cpp.edu> wrote:

David, All:

I have rarely been characterized as timid, David. Thank you. Another advantage of this book "baseline" would be that it could and should be built upon and its existence promotes and enables building upon.

For example, Hierarchy is represented by many more than Allen. There are nice published summaries by Wilson, Wilby, Salthe, Simon, Pattee, Miller, Bonner, Eldredge, Odum, Mesarovic, Weiss and more that are at least as penetrating and detailed (book length) as Allen. I myself have 27 products on Hierarchies. Existence of the baseline allows stable addition of extensions that need to be known. Also there are many more than 13 "traditions." In any case, what exists already -- the at-a-glance summary graphic you have produced --  is very useful.

Len

On Aug 26, 2015, at 9:35 AM, Richard Emerson <reme...@gmail.com> wrote:

Janet

I agree with Len and David, short expositions of the each of the entities would be a great book for introducing  systems science, guiding the study of it and suggesting further research in it.   

Your diagram is also a great framework for developing questions to ask.  A thought provoker.  I think you could keep an army of PHD students busy for years, and not only busy but productive in both understanding the development of the field and pushing the boundary.  For example, what other systems science concepts could be included.

Brilliant

Dick 

[Richard Emerson]

Janet

Perhaps you could take a leaf from the outsourcing, crowd-sourcing, mechanical Turk folks--parse the work out.  I wouldn't be surprised if Len, James, Jack that the rest of us have our own favorite mentors and would be able to say a few words.  come to think about it, it might have already been said on this and similar threads in this forum.

Regards (with a wink and a smile)

Dick

[Janet Singer]

Dick – That's a good idea, though getting the structure for the crowd-sourcing worked out is itself a challenge. But definitely worth doing, especially now that this could complement David and Jennifer's AKG framing.

Jack – Warfield is there, with connections to two he cites as influences. I have been particularly interested in people who explicitly acknowledged that they were 'hybridizing' technical and philosophical lines of thought. Wymore is definitely worth looking into.

Janet

[Steven Krane]

Interesting that Whitehead is such an active node in this picture.  I thought Whitehead was a subversive influence to the Entity-Relationship-Entity paradigm that pervades much system thinking and language.  I’m not sure anybody is ready to let go of that, or has an idea what to try next…

There’s a lot of smart people here.  Maybe y’all are doing that already.  I don’t know.

On Aug 25, 2015, at 10:34 PM, Janet Singer <janetm...@gmail.com> wrote:

Len,

That is indeed an awe-inspiring story.

In honor of 'that time of giants living their natural lives in networks' (as you eloquently put it) I am sharing a sketch of mine from 2012 showing some of the connection I had noted.

<Pragmatic Roots Singer 12-0904.jpg>

[Hillary Sillitto]

Jack

A process is what a system does, in response to stimuli if you will. So it's dynamic. It's simply the answer to "why did that effect happen?" 

You need to distinguish between what the system is (configuration, if you like), what it does (process), its observable effects, and the observed effects. Only the last of these four categories come into human awareness. Once the effect enters human awareness, the sequence set out in my previous post is triggered, so that humans progressively infer the nature of the observable effects, process, and configuration from multiple observations of observed effects, possibly in controlled conditions. The end result is conceptual models describing our best-guess description of the inferred configuration, inferred process, and why they cause the observed effects of the real world system. Once we find we can predict effects we start believing the model and underpinning assumptions. (This is how Physics has worked over many centuries; and our technology systems wouldn't work if Physics didn't work; so as a method it's OK.)

Your notion of configuration as the set of parts waiting to be stimulated to interact is perfectly compatible with what I intended to say.

e.g. Effect - broken window, stone lying on floor.

Process: boy throws stone towards window; Secondary processes: air resistance, gravitational force, stone hits window, window suffers brittle deformation.

Configuration: boy, stone, air, Earth, window, floor.

In the case of catalysis: the catalyst would be one of the interacting parts, so I see no problem with including catalysis in the process-effects concept. (By the way, I never said a system has to have a fixed or known boundary.)

As you said, "effects happen", whether authorised or not. A natural real-world system hasn't been designed, so the notion of "authorised stimuli" doesn't seem to make sense. A natural system does what it does, in the spirit of POSIWID. Any formulation that artificially excludes natural systems is unhelpful given the issues we now face. A formulation that allows us to learn from and curate natural systems along with man made systems is surely to be preferred.

or: if you wish to define system as that which responds to authorised stimuli, that is a different starting assumption, which will lead to a different worldview about systems; I suspect the worldview would be restricted to man made systems, so it might be very valuable in technology engineering, but wouldn't meet one of my goals which is to find a single systems paradigm that allows us to deal effectively with the full range of socio technical and environmental systems, and with both man made and "discovered" systems.

Hillary

[Janet Singer]

Steve,

I was also surprised at the extent of Whitehead's influence, since I started on this research 25 years ago thinking the reason the GST project had stalled was unawareness of the integrating potential of process-oriented foundations. 

The way I've come to see it is that the first development of systems thinking and language was largely compatible with the dominant views, i.e., it went as far as it could without being seriously disruptive. There was often push-back to even those modest innovations, but over the years the Entity-Relationship-Entity systems approach, as you characterize it, became commonplace and accepted.

Recently, however, developments in conventional science (mirror neurons, quorum sensing by bacteria, objective emergence and 'taking time seriously' in physics, pan-psychist proposals by neuroscientists, etc.) are opening – and actually leading – the way for a more sophisticated systems-oriented or emergentist or fundamentally complex paradigm. The full implications of the ideas of Whitehead, Peirce, Robert Rosen, etc., can now be more openly pursued than they could be during the early years when they were subversive.

I think we are seeing this across disciplines and application areas. I would cite this as the primary reason for optimism that the vision of systems-oriented unification of knowledge can actually be achieved. (Other reasons are 1. Growing recognition of the need for cross-disciplinary knowledge integration to address concrete problems, and 2. The promise of technology to support this 'collective actualization', as Duane Hybertson terms it.)

Janet

[Lenard Troncale]

Oh! Janet!

I added those names for Hierarchy just to show that once one has the stable framework, many others can add to parts of it such as more "traditions" (I love that you call them that) and more people for each tradition. Or people favorite to some groups as Jack just suggested for Warfield and Wymore. SO I think you are already following your own advice by presenting a framework. I do not think you were suggesting it was a final or finished, but rather something useful to use (which most others saw it as such).

Len

[Janet Singer]

Len, it's ok – the emoticon got deleted in the version I posted, but I know you weren't nitpicking.

I wouldn't have shared something so incomplete but for your inspiring story about Miller, but I'm glad I did because the response gave me food for thought. 

I'm also glad it sounds like you are making progress on your book.

Janet

[John Kineman]

Hillary, Jack,

Is it possible to include intentions and plans in the "process"?  If so, I think it is closer to the generalized process philosophy of Whitehead and similar views.

By 'authorized' I suppose is meant intended by the designer/builder/engineer. But a critical part of sustainability, and reliability as well, is response to unintended stimuli. I'm not sure what's being said here. Is it another way of talking about the system's purpose and design?

Configuration might make sense in common language, but be aware that it has a specific meaning in classical physics - it is the physical configuration and would not include the situation and effect of things like policies, usage habits, functional substitutions, re-design, stakeholder preferences, satisfaction, selection for a purpose, or anything contextual other than natural laws and perhaps some sensitivity analysis to determine stability of the configuration against perturbations in the configuration space. In classical physics, all that ever changes is configuration; all else being given in the big bang or not part of the consideration.

In relational theory, context is extremely important and effective. It is via context (formal cause) that a catalyst works - a catalyst can be an example or analogy for formal causation. Its an analogy if the catalyst operates by the same laws as the main process. Its an example in a slightly looser sense of 'catalyst' that operates by an incompatible formalism - for example agricultural policy regulating the climate system or enabling certain phenomena to occur that would not have occurred before. The process governing policy is not reducible to the process that may result in changed climate patterns, so they have to be coupled models and thus together form a complex with respect to that coupling.

Just so you know, Rosen's definition of "Structure" was "what a system IS" and "Function", "what a system DOES" -- both in a specific context. In more detail, what a system is would include its configuration, its composition, its design, and its purpose - any possible answer to a "Why <system>" question.

John

[Hillary Sillitto]

John

There is a large overlap in usage between functions and processes, and they are often virtually synonymous, with some cases where one word definitely fits better than the other.

I'll let Jack answer on configurations, it's his term.

Given that the point of the previous conversation was to identify whether systems could exist in the material world independent of human thought, it needs a bit of thought and care to work out where plans and policies - and planning and policy making - fit in. But in principle all the elements you mention have to fit somewhere. If intentions are a form of process, what is the effect or output created by intentions? Aah - but - Do intentions exist in the material world or only in human thought? Does human thought happen in the material world? Does it matter?

Hillary

[Jorg Largent]

Makes one wonder.  If a process serves no function, why does it exist?

---

Date: Thu, 27 Aug 2015 23:40:49 +0100

Subject: Re: [SysSciWG] Definition or Types of System

From: syss...@googlegroups.com
To: syss...@googlegroups.com

[Lenard Troncale]

All,

I am incredulous that humans remain so anthropomorphic or humanocentric that this long after proving the earth isn't the center of the universe, that it is round, that our galaxy isn't particularly special, that our solar system is just one among many, etc. we are still arguing over whether or not "systems could exist in the material world independent of human thought." Absolutely incredible.

Now whether or not the term "system" applies or all the artefacts of terminology in multiple languages and neural nets apply, is another question. But it has been shown by a vast array of mutually consistent timeline experiments in physics, astronomy, cosmology, geology, chemistry, biology etc. that we are totally dependent as a species of rather less than one million years of intelligent existence on these systems that were stable for huge amounts of time before we even became consciously aware of them....

.....that geological formations take tens of millions of years to achieve their current form,

.....that there are fossils as old as 2.5 billion years old,

.....that the earth and its solar system is 4.6 billion years old,

.....that the universe itself is just under 14 billion years old

and all the processes recognized fit into these ages, and still can be seen operating in their domains.

And somehow all this depends on humans to build models of them? Or correct their models of them with more experiments to modify in small ways what we humans have learned about these processes and timelines? Nonsense.

Natural systems exist and long existed before we began to become aware of them and even use faulty language to try to capture their reality. Deal with it or deny all of all of the natural sciences.

Len

[Lenard Troncale]

Jorg,

Nicely put. It is quite likely that some processes are min-max potent in a multi-dimensional space that we are not even aware of....  (meaning requires a minimum of energy, entities, events, time, etc. to achieve a maximum of utility to the sustainability of the defined system). That humans do not think in these terms is not the fault of natural systems, it is the fault of human limitations.

Len

John

Nice...

-- 
The SysSciWG wiki is athttps://sites.google.com/site/syssciwg/. 

 
Contributions to the discussion are licensed by authors under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.
--- 
You received this message because you are subscribed to the Google Groups "Sys Sci Discussion List" group.
To unsubscribe from this group and stop receiving emails from it, send an email tosyssciwg+u...@googlegroups.com.
Visit this group athttp://groups.google.com/group/syssciwg.

For more options, visithttps://groups.google.com/d/optout.

[Steve]

One could argue that insistence that "systems" exist in nature is humanocentric.

Nature exists. Although, the leading edge of physics today suggests that we are probably living in a hologram... in which case the earth, and everything else actually is flat (2-D)

I know it's troubling, but the problems with objectivist metaphysics are not going away.

How we categorize, including categorization of phenomena as system or not-system is an interesting study... to me anyway, maybe not to others. It seems fundamental, essential, and largely opaque, but, cognitive science is pulling a few threads. 

[Mike Dee]

Does Nature know it is a system?   Clearly nature exists, and nature does what it does (including what man does), but does nature know it is a system composed of many other systems?

Alternatively, is it not the observation of a phenomenon that enables the observer to define a "system" in terms of a model (right or wrong, without complete precision) describing the phenomenon?  

It seems that a "system" is a construct of human thought that describes some existing or wished-to-be-true relationship between things.

I think we are confounding the use of the word "system".  Perhaps the word "system" has contextual definitions?

MD

[Hillary Sillitto]

The interesting follow on question is whether there is a minimum definition of 'system' that is transferrable across contexts, and whether and to what extent any richer and fuller definitions are context specific.

Hillary

[joseph simpson]

Hillary:

Interesting idea and approach.

Mary and I use the following definition of a system:

"A system is a relationship mapped over a set of objects."

The minimum number of objects required for this system definition is determined by the relation type.  We require at least two objects (dyadic relationship) but there are other relationship types that require a different minimum number.

One interesting aspect of the word system is it is a collective noun. This aspect of the word system is indicated by the phrase "system of systems", or system of logic, or system of rules, or system of play.

Take care, be good to yourself and have fun,

Joe

[Lenard Troncale]

All,

I love these four recent responses to my admitted diatribe from my own past frustrations.

Why? Because I never deny that there is a huge and valuable set of insights that come from including humans as the observers and their undeniable effects on what we think (temporarily) is the reality. Logical Positivism, in its hopes and hubris, ignores humans and that is not useful. Like your responses I feel that cognitive science (and its friend philosophy) have many insights that they have already given us and will hopefully continue to give us.

I think it is a matter of balance. Denying or ignoring what we have achieved in understanding of the timelines and relative independence of aspects of real, natural systems is not balanced. For those who do not accept any reality outside of the human awareness, which is the operating position of many today in the anti-science and constructivist camps, it simply is not balanced. So definitions of systems should cite ALL of the camps and the strengths and weaknesses of each letting those of us who are students of systems to form our own synthesis and conclusion. The topic of systems in general especially is influenced immensely by human attempts to model with our minds what might be out there. But that is what experiment is all about.......trying to discipline our use of words and our neural net models with what might be out there. 

And in response to the expected citation of some of recent results showing that many "science" conclusions are inaccurate, please note that the most recent reports are on medicine and social science. I maintain that social science is not science (only because of the relative complexity of that studied)  and that medicine is studying the most complex of systems and has yet to incorporate the balancing of recognizing these as systems. I concede that the all important step of "reproducibility" in science is being currently ignored such that half of science studies show causations that "are not as strong" as reported (not false; just not as strong), I would say that there are textbooks full of past reproducible results that science is built on. That current work is weak does not condemn the whole of past science going back 400 years or more. Care to drop objects from a tower? Or compare such droppings to objects moving near the speed of light?

Ultimately I am just pointing out that folks working in human systems of all types, including SEs, are so focused on the relativity of human systems that they are missing the huge advantage of understanding systems that comes from studying and learning from self-defining systems that have existed for billions of years. I feel like a voice calling out in the wilderness.

Len