Computacional Exploration of the Structure of Visual Relationships

On this research, we are studying the methodology created by professor Rowena Reed Kostellow on the structure of visual relationships with the employment of computation design

On this research, we are studying the methodology created by professor Rowena Reed Kostellow on the structure of visual relationships with the employment of computation design. This methodology was developed over five decades and has been a model for design education over the world. It trains students and designers to be more sensitive to the abstract relationships of form and teaches a built-in system for form finding. It’s important to mention, that our idea here, is not to propose the use of computers instead of hands in Professor. Reed’s problems. On the contrary, as an enthusiast of this educational approach, our goal is to better understand this system. Hence, we are searching for hidden patterns in this system, that only computers can process.

To better understand the system, we defined a three-phase project. The phase one consists of the creation of an algorithm that virtually replicates models of one of the methodology’s system. On the second part of the research, we are going to create an analytic algorithm to calculate complex data about the compositions created in phase one. On phase three we hope to create a taxonomy of the models generated based on the information collected during phase two.

 

We based our first computational experiment on Rowena’s methodology first problem, called as rectilinear volumes. The objective of this exercise was to create a unity from three rectilinear forms as essentially different in character as possible. We used the exercises principles as rules for our generative algorithm. Here are some of those rules:

 

  • Volumes should vary in character
  • Differentiate the volumes by choosing a dominant, subdominant and subordinate forms.
  • Vary the inherent proportion, comparative proportions and overall proportions of the composition.
  • The axis of the volumes should be perpendicular to each other (in this exercise).
  • The design being aware of all sides of the piece.
  • Consider how the volumes are joined.
  • Dominant and subdominant volumes should be complementary.
  • The dominant form should be in the most prominent position.

To create rectilinear forms of varying proportions we created a system based on the Zigg diagram.

To create rectilinear forms of varying proportions we created a system based on the Zigg diagram. This chart was introduced by T. Zingg in 1935, that is used to plot the relative dimensions of the long, short, and intermediate axes of a particle, allowing its shape to be classified as bladed (see blade), oblate, equant, or prolate. With this diagram, after generating random rectilinear shapes, we can group forms with similar proportions for later cross combine them in the compositions.

The analyses we are developing have two types quantitative and qualitative.

In the end of her life on 1988 professor, Rowena was already concerned about the use of computer design in a field that the touch and sight are vital. In the book Elements of Design, author, Gail Greet Hannah states: “Computer-aided design was on the horizon by the end of Rowena Reed’s life, although it had yet to transform design practice in the way we now take for granted. She was especially concerned about the impact of the computer on the practice of three-dimensional design and cautioned against using the computer to do things that she believed only the human eye and hand could do. It now remains for those who value her approach to design education to make the intellectual leap and to devise new ways to integrate these new opportunities and new modes of expression with the traditional 2-D/ 3-D creative process.”  We believe that experiments like this one can indicate a way of integrating Rowena’s methodology with the use of computation and rapid prototyping without taking out the focus no the intimate relationship between human and shape.