June 15th, 2004

they_fly

On artificial languages "to speed up thinking"

Regarding this discussion on the "Speed of Thought"

1. What exactly are we to speed up? (a) thinking up new ideas (b) conveying to the reader the interplay between language patterns (e.g. poetry, literature) (c) recording and conveying reasoning (partly formalized already in math and logic).
Without defining precisely what it is we are seeking to perfect, we'll fail. Language IS TANTAMOUNT TO thinking in its non-visual part.

2. Thinking is patterning. Among the most useful patterns are similarity, contrast, metaphors, metonymy etc. Therefore not to limit thought the "better" language must not be based on logic alone

3. Natural languages are built on these mechanisms from (2). Therefore by synthesizing a new language from scratch we effectively dump thousands of years of thought and resulting compressed knowledge about the world.
We might think up new metaphorical understandings, sure enough, but we'd be kicked back into the stone age and would have to reinvent thousands upon thousands of wheels.
Whatever improvement we are to create, it must not dispose of the natural language. Rather, it should look more like a kind of slang optimising SOME aspects of the natural language (which, as we remember, IS thinking itself).

Let's now limit ourselves with pattern interplay and invention.
Which known techniques have people successfully used in advancing their knowledge?

(a) "chunking" and appropriate nominalization. Raising the level of generality by introducing new terms that substitute a lot of repeatable explanations. E.g. in physics: according to the second Newton Law bla-bla. No need to repeat the reasoning behind it.
Raising level by introducing and naming new entities (not named - not known) to compress expression of previous knowledge

(b) Revealing the structure. Logic and logic notation did just that. However as logic is not sufficient, we have to look into widening the application of this principle.
Good example is deBono "po" operator:
Birds fly south in autumn - po birds.
Or, using brackets and parentheses for grouping, we could transcribe it as:
(po Birds) fly south in autumn

So, we could introduce (an open) set of thinking operators, such as
gen - generalize; part(icularize); sim(ilar); syn(onyms); met(aphors); dec(onstruct into elements); def(ine new entity compressing above);
and so on.

For example:
gen [(po Birds) fly south in autumn ] -> migration;
gen[po(autumn)] -> other times;
part[last] -> hot, cold;
gen(last) -> change of environs --> {running from change}.

Result of the game (Остапово пособие для журналиста Улучшанского.):
*** трусливые осенние птицы бежали от холодов. ***


OR:
bridge connects two opposite (po river banks);
gen(last)-> island and bank; left side and fight side etc.
common (last) -> two pieces of flat land;
po(flat) -> negate(last) -> two landings on different levels;
--> tower as a bridge between earth and sky
analogy (last) -> Babilon tower.
result
Eiffel tower as a bridge between earth and sky, which God liked so much that He decided not to punish the builders for arrogance (дерзость). After finishing the Tower the French, to their own astonishment, found themselves still speaking the same language.
эйфелева башня как мост между небом и землей, который настолько понравился Богу, что он решил не наказывать строителей за дерзость. Закончив Башню, французы к своему удивлению обнаружили, что они по-прежнему говорят на одном языке.

4. How do we discover the elementary thought operators? - they have been discovered already, described really well as "invention techniques" and in structural analyses of natural languages. In fact, they have been described so well, that thinking techniques have been introduced into school curricula in some countries - and, as I heard with great success.