On misconceptions about separation principles. Part 1.

Y. B. Karasik,
Thoughts Guiding Systems Corp.,
Ottawa, Canada.

In a series of 1974-75 publications (beginning with "Mathematical Elements of the Theory of Heuristic Methods") I proposed and developed the idea of generic separation principles (in addition to separation in space and time proposed by Altshuller in 1973). Specifically I argued that any physical contradiction can be resolved not only by separating contradictory requirements between different locations in space or different points in time but between any different things1. For example, contradiction "X has to have property P (or has to be P) and has to not have property P (or has to be not P)" can be resolved by separating P and not P between T1 and T2: T1 is made to have/be P and T2 not.

These works led to the concept of separation principles in TRIZ. However, the meaning of separation very often gets lost on TRIZniks. Some people think that if something resolves contradiction then it is a separation principle. For example, employing a phase transition (as well as any other physical effect) may solve a problem and thereby resolve contradiction. But it does not make the physical effect a separation principle. It just implements separation prompted by a separation principle. Usually, this separation principle is separation in time or in space2 and the correct formulation of contradiction is "object has to be solid (or fluid, or gas, etc) and has to not be solid (or fluid, or gas)". Then it is resolved as follows: at the location A and/or at the moment X it is solid (or fluid, or gas) and at other locations/moments it is not. But when contradiction is formulated incorrectly it may indeed appear that the required separation is separation between different phases of the matter, which is a spurious separation. Just such incorrect formulations of contradictions led to calling a physical effect of phase transition separation principle.

When in 2000 I pointed this out in an article published in TRIZ-journal, I received the following response from Mark Barkan (addressed to the Editor of TRIZ-journal Ellen Domb):

Ellen Domb forwarded me Barkan's letter with the following request:

I replied:

Ellen Domb forwarded my reply to Mark Barkan with the following note:

A response from Barkan was not too long in coming:

At the same time Ellen Domb sent me a note indicating that my possible reply to Barkan would not be welcome: Needless to say that Ellen Domb did not publish the discussion in TRIZ-journal in spite of the promise.


1Actually not any but very specific ones. Others are just spurious and separation between them is never needed to solve a problem.
2Sometimes it is also separation between cause and effect, or between input and output, etc.