Mark Lazare 2005.01.20.2005
[From Rick Marken (2005.01.18.0820)]
RICK: Yes. But the aim of the reorganizing system is to organize a system so that it can keep error close to zero. The goal is zero even though the actual error is rarely actually zero.
Mark: The goal is zero… Sez you. But it is a program you wrote; it is not a model if it does not act and behave like a real person.
Mark: A real HPCT model would allow for a greater “acceptable error” “intrinsic error” or larger integration factors as one goes up the hierarchy. Just as do real people.
A living control system hierarchy starts with the 4 building blocks of DNA under very â€œtightâ€? control and then has ever increasing levels of control from the concrete to greater levels of abstraction.
The shorter amount of time a system takes to complete the feedback loop, the less error is inherent in the system. Conversely, the more time it takes to complete the feedback loop, the greater the inherent error in the system. A real model would address this issue as it relates to “acceptable error” “intrinsic error” or larger integration factors.
RICK: You could have the reference be non-zero but then what value do you choose? The ambient level of error in a control system depends on the gain of the control system.
Mark: I think the gain is a function of the integration factors
RICK: So a system with high gain may be able to keep the absolute level of error close to .001 error units. Such a system would actually be doing poorly if the ambient error went to .01. But .01 might reflect quite good control in a low gain system. From a modeling standpoint, it seems simplest to design the reorganizing system so that the reference for error in all systems is zero.
Mark: This is my point, the â€œsimplest to designâ€? approach to error you have taken is intellectually dishonest, and model ceases to be a model. It is like throwing a block of wood in a lake and calling it a model boat. Sure it floats, but in all other dimensions it fails to be a model.
Rick: I’ll answer some points now. You said:
As you go up the hierarchy you must build in greater level of Â³acceptable
error across a levelÂ² (X) or you would be in constant reorganization
Rick : This is not true. There is no concept of “acceptable error” in PCT and
certainly no requirement of a greater level of it as you go up the
Mark: That is a failing of or a shortcoming of PCT – that is what I am addressing.
Rick: In fact, the ambient level of error at higher levels of my spreadsheet hierarchy is actually less than that at the lower levels.
Mark: Which is again intellectually dishonest, and the spreadsheet hierarchy model ceases to be a model.
Rick: Acceptability of error implies some threshold level of error, above which reorganization starts. This is not the way reorganization is conceived of in PCT
Mark: That is exactly my point. â€œPeopleâ€? have “threshold levels” of error, both in a controlled process and along each level of control in hierarchy. Spreadsheets may not have threshold levels of error, but people do. Therefore the conception of reorganization in PCT needs work!!!
The model of reorganization in PCT assumes that the rate of reorganization (changes/unit time) is proportional to the size of the perceived error in the control system being reorganized.
I believe that is a false assumption. There are so ways to prove that statement false in just every day experiences. For example the sudden death of a spouse or child = large perceived error. According to your statement and your spreadsheet a person would quickly get over that loss because a control system would quickly compensate for such a great and sudden error. Additionally a 20-year alcoholic wakes up after 4 failed marriages, the loss of many friends and untold number of jobs and says I am going to stop drinking today (again), but this time he actually does.
My point is PCT does not have a working model of reorganization, awareness or attention.
I humbly think my way of looking at reorganization more closely resembles the way people deal with error as a part of reorganization, awareness and attention.
Mark A. Lazare, Managing Partner
Compass Mental Health, LLC
4500 N. 32nd Street, Suite 104
Phoenix, AZ 85018
HIPAA Confidentiality Notice
The documents inside this electronic transmission contain confidential information belonging to the sender that is legally privileged. This information is intended only for the use of the individual or entity named above. The authorized recipient of this information is prohibited from disclosing this information to any other party and is required to destroy the information after its stated need has been fulfilled, unless otherwise required by law.
If you are not the intended recipient, you are hereby notified that any disclosure, copying, distribution, or action taken in reliance on the contents of these documents is strictly prohibited. If you received this electronic transmission in error, please notify the sender immediately to arrange for return.