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

A few philosophical thoughts (on Rosennean models and modeling in general)



I've been musing on the concepts of "reality and reasonable facsimile" or natural systems and models of natural systems-- and what happens when the two are not in "congruence". Such musings lead naturally into thinking about new ways of establishing or restoring congruence between natural systems and the models we create of them...
 
One of the things that I notice in science is that scientists often become engrossed in "processes"-- as do people in any profession, I suppose. It happens even in non-professional areas of life, as well (I remember a personal trainer, a few years back, exhorting me; "You gotta learn to love the process, Jude...") I suspect this tendency is endemic simply because it is another aspect of human nature. However, the danger in this tendency is to become so engrossed in a process that the reason for the development of the process gets lost. I think this is what has happened in science generally-- science itself is "a process", after all. My father's trouble with mainstream science (and mainstream scientists) tended to be created from this attachment that human beings form to a process they like or feel comfortable with. Some scientists, particularly experimentalists, have actually specialized in "processing" and have little interest in the deep underlying issues. It's little wonder, then, that such folks didn't take very kindly to my father coming along and saying that reductionist approaches won't achieve answers to any of the fundamental questions in biology.
 
His willingness to sacrifice established "process" in order to pursue the original question or problem was one of his strengths. He also was able to keep his focus rather than be sidetracked by the myriad side issues that cropped up along the way-- and was occasionally criticized for not developing those side issues further. One of the side issues is "how to build models based on Rosennean Complexity Theory".
 
One of the things I've been seeing in several recent discussions on the list lately is a great deal of attachment to, or excitement about, various modes of modeling (modeling processes). The "best" modeling process for some complex system is only "best" in a context-- whatever that context may be. A change in context may require a change in process, among other things. Some spectacularly successful mode or process for modeling some complex system may only be successful for a short time. The first question, then, in developing any Rosennean model for some complex system is to inquire into the context-- or so it would seem from where I'm standing. Everything from that point on is generated because of the contextual influence and it seems very important to bear that in mind when looking at other applications for the modeling process that is created or when looking at remedial tweaking on a process that isn't working optimally.
 
The difficulty with such realities as human nature impinging on professional pursuits is that we tend to become very good at something, and then we get a lot of positive reinforcement from being good at it. It is human nature to marry our "traditions". When a scientific process becomes more important than the initial reasons for developing it, the situation is unbalanced.  In that case, if the original context changes too much for the "process" to be useful anymore, one would either have to take that expertise (in applying and/or manipulating the initial "process") to a new context which closely matched the original, or work to keep the context from changing. Ever.
 
Such areas of thought become even more intriguing when applied to the internal models that all living systems have as part of their complex organization.....
 
Ah, well... time to go do yard work.
 
Judith