Discussion                         Discussie  

In the examples the most straightforward working of the principle is shown. We can call this a rather stupid or time consuming approach. The interesting thing, however, is that all kind of decision-rules to improve its working into a more intelligent, strategic principle can be build in. Some examples of decision-rules are:
  • if partial structures are confirmed then these are not further differentiated. Only the part that is denied is differentiated (see 3.5.1 to 3.5.6).
This rule shows us two extremes. On the one hand a structure which is differentiated in its totality, so if one part is denied, the total structure in all its developed parts is differentiated. On the other hand all parts of a developed structure
are going to live a life of their own. So no relation with some total structure is maintained. The relation between total and partial structure in the first extreme we could call order in the latter chaos. Between these extremes there lies the possibility of organization, which is a field of free-choice. Within this field there is no one solution to a problem. It will always be a solution, other solutions stay possible dependent on the sofar developed means. When we look at fig. 3.5.1 to 3.5.6 we see that in the given example the right part of the structure is fixated. Only the left part of the structure is developed further. It goes without saying that the right part can be developed further at the same time. This development of partial structures at the same time is what is usually called parallel-processing.
Another rule could be:
  • Compare after an expansion both outside lines and if these are very different (given some criterion) then differentiate the largest one and expand.
  • Or compare after an expansion left and right lines and if they differ considerably then adjust… and so on.                                                                                           
A second feature is the possibility to describe the contour by means of the so far developed structure. This opens the possibilitv of pattern-recognition. All kind of calculations on the realized structure can be performed. So information can be delivered about size, edginess, symmetry, position, roundness and so on. Furthermore different descriptions can be very coherent or incoherent independent on the chosen contour or startingpoint.

This leads to the problem of ambiguity.

Six possibilities cover the field of this problem and within these we can describe three forms of ambiguity.
1.given the same contour and given
  • the same startingpoint, we can be confronted with the same and different descriptions. The different descriptions create the first form of ambiguity.
  • different startingpoints, we can meet the same or different descriptions. The different descriptions create the second form of ambiguity.
2.given different contours, there can be the same and different descriptions. In this case the same(!) descriptions create the third form of ambiguity.

These three forms of ambiguity deliver no problem. Differences in interpretation can be solved by either further development of one of the two structures on which the description was based or by adjusting the differentiationlevel (increasing the number of points to be realized on the contour) and/or increasing the number of descriptive categories. lts more a matter of effort than of insolubility.

 Volgende                          Contact                                Terug