WORKSHOP: LIMITS, CHOICE AND EXPRESSION




   I. INTRODUCTION:




        The starting point for the workshop is the question:

            "How is it possible to express oneself through the medium
             of the computer considering that, in principle, all possible
             pictures are potentially defined by the number of points on
             the screen and the number of colours available?".


        One may consider the following related points:


         i.   The possible similarity to human language systems where a
              large but still limited range of words seemingly allow
              infinite expressions of meaning.

              This invites the following questions:

                a.  How are can a limited repertoire be expanded to cover
                    a wide range of meanings?
                   (i.e. the number of possible statements be maximalized)

                b.  How important is the interpretation process?
                   (i.e. does the manipulation of meaning increase the
                    number of possible statements?)

                c.  Is verbal language a valuable model worthy of further
                    study?
              

         ii.  The possibility that all systems are in some ways limited
              and the limitations of the computer may be similar to those
              of traditional media like painting (where one is also limited
              by canvas size, palette, size and shape of brush stroke,
              art history etc..)

              Certainly it is possible to categorize different types of
              limitations found in daily life:

                a.  Physical Limitations:
                      i.e. some variations are physically impossible.

                b.  Conceptual Limitations:
                      i.e. some variations are possible but we haven't
                           thought of them yet.

                c.  Perceptual Limitations:
                      i.e. many variations are possible but we can't
                           distinguish them from others.

                d.  Moral or Aesthetic Limitations:
                      i.e. some variations are possible but we choose not
                           to use them.


                e.  Logical Limitations:
                      i.e. any variation which if it were possible would
                           contradict our system of belief and understanding
                           about the world around us.
                      i.e. variations which may be impossible within our
                           belief structure but may be possible outside it.

              NOTE: With the exception of Moral/Aesthetic Limitations,
                    which may be consciously made, it is very difficult to
                    determine how valid the limitations are.

                    i.e. We are surprised when an impossible variation
                         is proved (by experience) to be possible, but it is
                         virtually impossible to know if a variation is
                         possible or not before it is proved to be possible.





         iii.   The different (but equally valid) aesthetics of sterile,
                accurate and controllable digital systems compared to messy,
                inaccurate and relatively uncontrollable analogue systems.

                i.e. Analogue system: continuous space
                                      �like a topographical map where
                                       valleys gradually become mountains.
                                      �variation is implicit and must
                                       explicitly removed if undesirable.

                     Digital  system: discontinuous space
                                      �like a political map where
                                       countries have abrupt borders.
                                      �variation is absent and must be
                                       explicitly included if desired.
                                       (within limits of resolution)



         iv.   The difference between physical and conceptual actions
               (i.e. the absence of a perceived physical action changing
               the image may force one into more abstract modes of thought)





   II. TOWARDS SOME THEORETICAL SOLUTIONS:



        i.  Combinations and Grammars

             a. Combinations in Time and Space

                   A easy way of increasing the number of possible
                   statements that can be made with a limited repertoire
                   is to copy the basic elements and to re�combine them
                   in Time and/or Space.

                   i.e. the images may be limited but their behavior
                        may be complex and varied in time.
                       (morse code, digital sound sampling, etc.)

                   i.e. the elements may be limited but they can be
                        repeated in complex patterns in n�dimensional space.
                       (parallel port, colour bit�planes, Weaving, etc.)

               NOTE: The Space is often limited (i.e. the number of bits
                     in a computer system, the number of pixels in a screen)
                     so the patterns remain restricted by the number and
                     length of the dimensions of the space and the number of
                     basic elements.

                     One assumes that Time is infinite and therefore there
                     are an infinite number of patterns in time that can be
                     constructed even from a minimal repertoire of two
                     basic elements. Practical and Perceptual Limitations
                     may still be valid here too.


             b. Perceptual and Conceptual Categories

                   Although we can mathematically calculate the number of
                   patterns that can be generated by any given number of
                   repetitions of a basic repertoire, there may be a
                   considerable difference between the theoretical number of
                   patterns possible and the number of patterns actually
                   usable.

                   Perceptual Categories:

                     As noted earlier �some patterns may be indistinguishable
                     from others and the number of effective patterns reduced.
                     (i.e. is that line 36.5 or 36.6 cm. long?
                           is that note 366  or 367 hertz?)


                   Conceptual Categories:

                     Some patterns will form conceptual relationships with
                     other patterns.
                     (i.e. Some patterns may be rotations, mirrorings, or
                          colour isomorphisms of each other)

                     Depending on the circumstances the conceptual
                     categories may increase or reduce perceived differences.
         
               

             c. Filters and Grammars

                   For practical or Aesthetic reasons, some combinations
                   of basic elements or processes may be unacceptable.
                   (i.e. a large box can't be placed inside a smaller box,
                         Certain colour combinations may be unpleasant, etc.)

                   The rules determining which combinations are acceptable
                   and which are not are generally (in linguistic terms)
                   known as "Grammars". In order to work Grammars usually
                   require some kind of "Grammatical Categorization" of
                   the basic elements (or "Primitives") in order to specify
                   those categories of element which may be combined and
                   those which may not.
                   (i.e. a sentence requires a noun-phrase and a verb-phrase)
                   A grammatically correct combination of symbols is termed
                   "A Well Formed Formula". Traditional Linguistic Grammars
                   can decide if a sentence in a given language is correct
                   or not but are not very useful in the actual construction
                   of the sentence.


                   Conditional Filters can be used to remove unwanted
                   combinations of elements.
                   (i.e. Make Combination �IF Combination NOT Satisfactory
                        THEN Reject Combination AND Make New Combination)

                   Unfortunately the rules to reject unwanted combinations
                   may be difficult and sometimes complex.


                   Procedural Grammars can be considered as a set of
                   switches structuring the performance of a set of
                   procedures. i.e.

                      EITHER Modify size
                                   (case a) increase height
                                   (case b) increase width
                                   (case c) increase both
                                   (case d) increase neither
                      OR     Move position
                                   (case a) move x
                                   (case b) move y
                                   (case c) move both
                                   (case d) move neither
                      OR     Change Colour


                   Transformational Grammars are sets of transformation
                   rules which are so defined that they always generate
                   Well Formed Formulas (WFF's).
                  (i.e. Any combination of symbols generated by a specified
                   grammar are termed WFF's of that grammar, so one has to
                   define the grammar so that it can only generate acceptable
                   combinations)


                   Transformational Grammars obviously consist of a set of
                   transformational rules and look something like this:
                           x   �> xx
                           xxx �> yxx


                   This Grammar would generate a strings like this

                          x
                          xx
                          xxxx
                          yyxx  or xyxx

                   depending on it being read from the left or the right
                   (which would normally be implicit in the system).

                   Transformational Grammars are generally used for
                   artificial languages and have not yet been able to
                   generate all the grammatically correct sentences of
                   a natural language (i.e. English, French, etc.) and
                   it may be impossible for them to do so. They are easy
                   to implement in the "LISP" programming language.

               

        ii.  Organization and Interpretation
              (i.e. The creation and definition of Space)

                   The concept of variation implies that there is something
                   that can be varied.
                   (i.e. a "Control Knoop" that can be turned to increase or
                   decrease something, such as height, volume, speed, etc.)

                   So in order to experiment with combinations and/or
                   variations of basic elements we shall need to define
                   our basic elements and decide what we can do with them.

                     i.e. do we have simply a set of images?
                             (select image a, b, or c)
                          do we have a set of elements and locations?
                             (set element x at location y)
                          do we have a set of procedures?
                             (change colour, size or position of figure)
              

                   Obviously, the way we select or define our basic elements
                   has implications for what we can do with them. Our basic
                   decisions in fact define a "Universe (or Space) of
                   Possibilities". It may be interesting to observe closely
                   how our decisions affect the dimensions (or "Parameters")
                   of the space we are defining.




   III. TOWARDS SOME PRACTICAL EXPERIMENTS:




        i.  An Introduction To Computer Programming:


        ii. Some Practical Exercises:

               -enumerating the repertoire (coding?)
               -random generation of image
               -defining some basic repertoires and their variations
                   (parameters of space)
                   (alternative definitions)
               -controlling the variations
                    -random choices
                    -random clocks
                    -deterministic clocks (functions)
                    -feedback and self controlling systems
                    -interactive systems




   IV. TOWARDS SOME THEORETICAL MODELS:



     The Dialogue:

         It is apparently easier for humans to modify their external environment than their own internal physiological system.

                                                                                                                                                                              (Electron)

         Perhaps "Art" is "one" way of manipulating the external environment
         in order to modify our internal operations as a result of their
         reaction to the environment.


     Information:

         What is "information"?

         Are there different categories of "information"?

         What is the relationship between directly experienced
         "source information" and  "synthetic information" which
         is derived by some kind of processing? (Jonah)

         Is not all "information" the result of some form of processing?




   V. THE CONCRETE RESULT:



        An introduction to programming in machine language, given by one
        of the students, gave a practical insight into how interpreting
        binary codes as operating instructions enables a limited basic
        repertoire to be bootstrapped into complex behavior patterns.

        One may strongly suspect that it is indeed the interpretation
        process, by encouraging shifts into different "domains of
        discourse" which generally permits the basic repertoire to
        transcend itself.


        Although the students worked hard to master Basic programming
        techniques, only one group produced interesting results based
        on variation of the circle.

        Again it was interesting to note that although the program was
        based on drawing concentric circles (at different locations)
        the circle parameters quickly became subservient to those of
        the controlling loop structure (i.e. Increase or Decrease of
        radius, size of Increase/Decrease and colour of circle). This
        would appear to demonstrate once again how creating a "higher
        level of language" transcends the original limits of the basic
        repertoire �obviously not in a technical sense (of physically
        extending the repertoire) but in an organizational sense of
        introducing new ways of ordering the original material).


        The students who remained until the end of the workshop were
        generally more interested in their own explorations (which in
        some cases can be expected to continue beyond the workshop)
        than in the limited theoretical starting point, however it did
        prove to be a useful introduction in many ways.

        The student who posed the original question was too confused by
        technical practicalities to develop the theme further within the
        workshop. However she has promised to digest the experience and
        to present her thoughts at a later date.




                                                       Trevor Batten
                                                   Amsterdam, March 1995