+-------- 1 ------- 1.1 -----  1.11
   Welt Fall   Welt Gesamtheit Tatsachen Dinge  Welt Tatsachen Tatsachen
          |         +--------  1.12
          |         |          Gesamtheit Tatsachen Fall
          |         +--------  1.13
          |                                        Tatsachen Raum Welt
          +-------- 1.2 -----  1.21
                Welt Tatsachen       Fall Fall

although the coding of words by ascii codes doesn't contradict the arbitraryness of the langugage, it makes various things unreadable.

Simplified rules: because of this the coding scheme should be only used for concret objects. one removes the predicate, since it cannot be recongnized by an algorithm. also the subject is not emphasized anymore, all the arguments have become equally important. also the logical operations are removed. what remains of each sentence is the array of arguments, which are coded in proportion. when given the occurance of an argument, which has been used in one of the previous sentences, it will be replaced with the entire new sentence. but giving a kind of mutating recursion

#VRML V1.0 ascii
Separator {
Translation { translation 23 5 12 }
Cube { 	width 23 
		height 5 
		depth 12 }}
Separator {
Translation { translation 6 1 12 }
Cube { 	width 6 
		height 1 
		depth 12 }}

sentence 1.1:

#VRML V1.0 ascii
Separator {
Translation { translation 23 5 12 }
Cube { 	width 23 
		height 5 
		depth 12 }}
Separator {
Translation { translation 7 5 19 }
Cube { 	width 7 
		height 5 
		depth 19 }}
Separator {
Translation { translation 20 1 20 }
Cube { 	width 20 
		height 1 
		depth 20 }}
Separator {
Translation { translation 4 9 14 }
Cube { 	width 4 
		height 9 
		depth 14 }}

In sentence 1.1, the argument "welt" occurs again, so the Cube { width 23 height 5 depth 12 } object in sentence 1 is replaced by the entire sentence 1.1
following that we get:

1	1.1
1 + 1.1	1.1.1
1 + 1.1 + 1.1.1

Using that method, the result is a simplified representation of the text, still retaining the correct sequence of the propositions.