TRIZ RING: should or should not TRIZ become a science ?
(continued from PART 1 presented in Vol. 1, No. 12)

GF: We are not talking about a scientific foundation for the art of invention, the art that existed before TRIZ and still largely exists out of TRIZ (and which TRIZ dismisses as mindless "trial-and-error"), are we? We can discuss that latter, when we're done with TRIZ... Now, we are talking about a scientific foundation for this new thing, TRIZ. The objectives of both are the same - making inventions. There are two possible relations between them:
1) TRIZ is an alternative to the old art of invention, which does better (that's what TRIZ claims); I agree that having a scientific foundation probably is the only way to objectively do better;
2) TRIZ is a didactic/heuristic system of a generalized, concentrated and organized collective experience of the prior art of invention, which puts it into an efficient use (TRIZ claims this too!); no scientific foundation is necessary in this case.

All I see when examining elements of TRIZ - from the "inventive tricks" to the "standards" to the "multi-screen scheme" - is didactic/heuristic methods. These "inventor's instruments" are themselves inventions - Altshuller was quite good in this,- and they don't need any foundation other than explaining how they work.

Please, allow me to bring in a new analogy, just to demonstrate what I mean. Among other things, I teach a scuba diving course "Marine Life Identification". The course provides for, and I invent more of, various ways to classify and organize hundreds of fishes and other creatures for divers to be able to recognize what they see. These are more or less effective didactic/heuristic methods. They don't need a scientific foundation.

YK: If I understood you correctly, you distinguish between a scientific foundation for the art of invention and a scientific foundation for TRIZ. You agree that putting a scientific foundation under the art of invention will make it better. Better in what sense ? In a sense that it will make it independent of uncontrollable factors such as insights, luck, etc. ? In a sense that inventors would have a guidance in circumstances where previously they had to rely on insights and/or luck only ?

GF: The distinction I make is even more specific: I distinct between
1) pre-TRIZ art of invention, and
2) TRIZ, which is either
2.1) an alternative to the pre-TRIZ art, or
2.2) an addition to it.

In case 1: putting a scientific foundation under the pre-TRIZ art would make it... well, scientifically founded. Anyway, this is a separate topic, which might need a discussion, later.

In case 2.1: if TRIZ provides for a new way of making inventions, if TRIZ is to replace the pre-TRIZ art with a different methodology, by which inventions are determined rather than guessed, then, yes! there should be a scientific foundation for such "calculus". Do you think, this is the case?

In case 2.2: if TRIZ provides inventors with saw, hammer and even laser to do what pre-TRIZ inventors did with bare hands, it still operates in the same realm, has the same nature as the pre-TRIZ, and its foundations are not different from foundations of pre-TRIZ. It is the same work done more efficiently (e.g., identifying fish in 5 steps instead of random looking through 2000 pages of a fish species catalog). This case actually brings us back to case 1. In this case, TRIZ needs no more foundation than FCA, morphological analysis, multi-screen schemes ... oops, but it is TRIZ!?

YK: Thus, you distinguish between "calculus" (or algorithm) of invention, old fashioned pre-TRIZ technology of invention, and TRIZ. You agree that TRIZ is more efficient technology than the old fashioned one, but far from being "calculus".

Since TRIZ evolves, improves all the time, what does preclude it from evolving into a real "calculus" (or algorithm) at some point in time ? Why there should be some limit to its efficiency, falling short of the efficiency of algorithm ?

Can you propose a proof of impossibility of evolving TRIZ into a "calculus" (or a real algorithm) other than labeling TRIZ "heuristic", "didactic", or whatever system ?

GF: I cannot prove that by CHANGING (sic) TRIZ, FCA, morphological analysis, brain-storming or anything else, one cannot build a real calculus of invention (not necessary an algorithm, by the way). Moreover, I don't think so!

The key word in your question is EVOLVING, isn't it? What does it mean? I can think of two meanings:

1. A sequence of changes. After enough changes one can get something completely different, which has absolutely nothing in common with an origin. In this case, the result may be anything, calculus or not, and the fact that it EVOLVED from TRIZ would have only an historical significance. By making enough changes one can arrive from anything to anything, no doubt.

2. A framework of changes, i.e. keeping some core unchanged. In other words, the calculus would have the same basic principles as TRIZ. I still cannot prove it's impossible. I just see that all basic principles of TRIZ are purely instrumental, not conceptual or fundamental.

Here is what I think: if there will be a calculus of invention, it will have no common principles with TRIZ of today. This TRIZ might or might not somehow help with building this calculus, like (here it comes - analogy) alchemy might or might not somehow help with building a scientific chemistry. Can you point to the principles, which will hold, in your opinion?

YK: From your answers, it looks like the difference in our positions stems from the difference in interpretation of word "TRIZ". For you it is something limited to its present contents. For me it is some new area of research that aims at formalizing inventioneering, at creating an algorithm (or "calculus") of invention (and creativity in general).

I do not know what will have preserved from the current principles of TRIZ when the "calculus" will have eventually been created (if ever). However, one should not be misled by the word "theory" behind acronym "TRIZ". It has never been a static theory, even in Altshuller's days. Even he constantly changed it trying to improve, make more efficient. From what I have been observing, it is rather a new area of research than a theory. But the acronym stuck, and we continue to label the area "TRIZ".

For the time being, I do not see any reason why attempts at creating "calculus" (or a real algorithm) should fail. But at the moment, you are right: the current TRIZ does not meet the objective set by Altshuller, namely, create an algorithm of invention.

The current TRIZ is like a glass just half full of water. That is why one can also view it as (half) empty. Neither view is wrong.

Thank you for the discussion.

GF: You're absolutely right: we interpreted word "TRIZ" differently. It was not clear in the beginning of the discussion. The discussion is productive, just because this clarification is important. Thank you, too.