Hi Dear Rai

I think your question is interesting one, I have red little about that. Garud, R., & Kumaraswamy, A. (1995). have writen research paper about  " Technological and organizational designs for realizing economies of substitution" . [Strategic management journal, 16(S1), 93-109.] and

Jayal, A. D., Badurdeen, F., Dillon, O. W., & Jawahir, I. S. (2010).  have examined  " Sustainable manufacturing: Modeling and optimization challenges at the product, process and system levels" .[CIRP Journal of Manufacturing Science and Technology, 2(3), 144-152] .

I am sending links of both papers herewith as an attachments. I think you can find out some important ideas from them. 

https://archive.nyu.edu/jspui/bitstream/2451/14190/1/IS-97-32.pdf

http://www.sciencedirect.com/science/article/pii/S1755581710000131

regards

Dr. Kumara

I think that this is an interesting and important question(s).

However, coming at your question from a mathematical/computational perspective (sorry! -- it's what I do!), my immediate thought is that all the elements of it (eg "complexity of a system", "compromise" ,"integrity", "emergent behaviour", AND the possible operations on the system that might reduce its complexity) must be given much more precise definitions than they are normally given,  if progress is to be made -- indeed they have to be defined in mathematical/computational terms for there to be a well-defined problem that can then be concretely solved. Otherwise there is a very real risk that ambiguity dominates?

JD

Dear Veerendra K Rai

a good question and the answer depends on your philosophical stance. There are researchers that assume they are unable to fully capture reality and consequently any description of a system will thus lack a full coverage.

It is argued that models are mere aproximations of the systems we research upon. Of course there are also researchers that argue these models to be imperfect but acceptable substitutes and thus can be observed and studied easier.

Kind Regards

Bing

I do agree with James Doran on mathematically defining the terms used in the question. I request him to take the initiative and define them based on his understanding of these terms.

The following line of research seems to be one way  (of many ways!) to begin to define more precisely the terms in the original question.

In my MoHAT project  a complex system is defined with computational precision and then the question arises whether the system can be made less complex whilst preserving its behavior.

More exactly, in MoHAT I have defined, specified and implemented on a computer (in the programming language C) an interesting type of complex system that I call a MABEEA system 

At the heart of a MABEEA system is a (potentially very large) set of precisely and simply structured IF..THEN… rules. These rules match against and write values of the MABEEA system's state variables.

Subsets of these rules and associated variables can sometimes be “read” as computational AGENTS, perhaps forming some kind of SOCIAL COLLECTIVE, if that interpretation is under investigation.

As a particular MABEEA system runs (typically from a randomly chosen initial state) its “history”, H say, is generated and recorded. Thus a natural question arises: can the rule set R that generated history H be made less complex (e.g. with fewer rules? differently structured rules?) yet still generate history H?

For more detail on MABEEA systems, including the computer programme,. see recent MoHAT project updates here on RG

Best regards

James

Thanks James. Good Point. Part of my project I am currently involved with is about "What IF" analysis in IT service operations management domain. It did not occur to me if I could look at "What IF" from that perspective. Now I will.

Possibly, each system has to be looked at individually.

One related question. How would we know that in the process of reducing complexity systems integrity has not been compromised? Can we say that if pre and post "complexity reduction" system's behavior remains the same- whatever metrics we use to measure behavior- then system integrity is preserved?

Can we say that if pre and post "complexity reduction" system's behavior remains the same- whatever metrics we use to measure behavior- then system integrity is preserved?

My answer to this question  is "yes" ,  but obviously it depends what exactly is meant by "integrity" and I suspect  that some would argue that integrity is not just a matter of behavior? And the behavior examined has to be representative, substantial etc?

JD

Yes James. It raises further questions. How we define "integrity" and whether integrity is more than behavior? Are there instances when "integrity" is behavior and when "integrity" is more than behavior?

In System Dynamics structure and behavior of a system are closely related. Structure determines the behavior, but they are not the same. Structure is made of feedback loops in system dynamics.