This research starts from the intention of automating the composition of drawings through programming languages. The word “design”, present in the title of the proposal, refers to both to design as planning and as the action of drawing.1 Drawing was chosen as a starting point because it is a language used in several activities comprehended between the domains of art and science. It is an artistic proposal built from digitized images whose visualization, combination, and organization occurs through algorithms. “Adrift” refers to the drawings developed in this proposal and their characteristic of altering through time, going from incomplete to completeness; they are compositions that change and this is the reason they are so different throughout time, complete at each instant.

The research was developed in artistic and theoretical fields. In the first field, it follows a path that starts in the technical apprehension of the resources and syntax of a programming language and goes up to the development of experiments that use it to organize fragments of image in compositions. The theoretical research approaches the relationship between formal rules and artistic creation, not necessarily with the use of machines, technology or programming languages. An issue that often came up during this research was the confrontation between formal programming rules and the sensitive content of the drawing. The motivation for the use of formal rules is the possibility of automating processes. The idea that one given composition is as possible as several others that were not done is in the beginning of this proposal. Programming serves the objective of exhibiting several drawings resulting from determined concepts. This expansion is favored by the programming through its iterative and recursive models, both formal generative processes executed in cycles.

In this proposal, several compositions that are possible to be created are grouped inside a specific operation, which dislocates the creation process of the image from the action of tracing it to the definition of the operation that will generate it. As we draw using traditional materials, the alterations that occur in the medium are consequences of choices.  These are decisions that take in consideration local or global aspects (in relation to the space of the composition), of an objective or subjective character, emerged before or after the action of drawing. For a programmed image, the choices referring to its creation are made before or during the action of programming. The changes in the composition that originate from the execution of the program are previously determined and are not changes in the support, but visualizations presented by the program while it is executed. The outputs shown are possible and pre-determined multiple compositions. Their exhibition does not involve creation anymore, but execution of the code. In this context we see the difference between drawing and design. In this proposal, design exists in both senses: in the language that defines the operation of the compositions and in the images presented by this operation. Images are generated by an operation programmed through computational models that are presented as structures of formal organization.

The programmed mediation aims specifically at returning programmed answers conditioned to determinist rules. The presence of determinism in computational programs does not make their answers so predictable or regular because a program usually does not execute only one function, but many. Programs are structured as systems in which several parts compose an organized whole. The use of systems in any phase of construction of an artistic object characterizes the practices gathered around the concept of generative art. According to the thought developed by Philip Galanter, generative art is as old as art itself, so it is not necessarily built with digital tools, but by the automation allowed by systems of algorithms executed by the available or chosen forms. For this work, the programming language used was that of Processing, a language in open code that was developed with the intention of presenting concepts of programming within a visual context. Processing was thought as a learning, prototyping, and production tool to be used by students, artists, and designers. In this context, basic resources can be assimilated or more complex programs can be built in order to make drawings, interactivity, and animation.

Because they are systems, we can consider the execution of generative programs as throwing of dices, only many of them at the same time. For this research, the first compositions controlled by programming were made with the random function, which returns apparent random values. These values can control any parameter of any object and, in the composition of this drawing, the function was used to control de location of the images that appear at each new execution cycle of the program.

The objective of the random function is to simulate the casual choice of values, but this is done in a very ordered way, which is why the generated values are, in fact, pseudo-random. Let us see how it works: a random function programmed to return values between one and six will return, in 60 tries, approximately 10 times each value. The algorithm does not allow us to predict which the next value will be, but we do know that it will one of the predicted ones and certainly the values will come out in the same proportion. This is one of the ways to create a model to simulate a situation perceived as being somewhat random, but generating an ordered, and not a casual, result. We can make number 6 have 50% chance of coming out. In this case, the random function will always act as loaded dice. In a composition that only uses the random function to exhibit the elements in the two-dimensional space, they will show evenly on the screen. Probability can be used to make the elements accumulate in certain areas of the two-dimensional space first, but with time the whole screen will be filled uniformly.

In this first experiment, there is only the interference of probability applied to the random function. It is an extremely simple and deterministic program. It is possible to make random values interfere in other values, which brings the program closer to the idea of a system in which several parts work as a group. We can also program it in a way as to vary the location of the image fragments according to the position of the mouse and add to that some pseudo-random results. When the mouse is activated at the bottom left corner of the screen, it can make the image fragments appear in any other place that is not the bottom left corner. Nothing prevents the interface to be programmed in order to react in unusual ways, “usual” here meaning the simulation of reality, i.e. an “expected” functioning, a “proper functioning”. The programming language does not have any energetic connection with reality; the reality of the language is its own syntax and operation.

In the following experiments, we used images that combined in square or hexagonal patterns. These images fit into a grid and there is an overlapping of progressively smaller grids in which the lines of some intersect with the lines of others. The intention now is to advance in direction of more complex compositions in which local interferences can alter the composition of a certain area of space. It is possible to make the presence of a determined fragment affect another according to some preconceived meaning. Only forcibly are the formal structures of programming capable of working with meaning. Programs do not work with meaning. It is us that eventually perceive meaning in the execution of programs. Art that works with techniques of programming stays in this limit between pure order and the deviations to the disorder that characterize the aesthetic experience.

In this sense, there are algorithms or forms of using them that provide non-uniform organizations, closer to our perception of organic phenomena. These algorithms are capable of producing complex operations due to the intersection of simple rules. One of the concepts involved in the concept of complexity is precisely about bottom-up organizations, i.e. complex totalities derived from rules and simple elements. In programming language we can make this time of organization emerge when there is some kind of local rule to which elements mutually react at the presence of other elements. This behavior is called self-organization. Self-organization can play with determinism of the languages in the sense of not turning the predictable results to the perception and even to generate behaviors that were not programmed, but observed subsequently. An example of self-organization in programming is the “Game of life”, developed by John Conway. This program consists of a grid in which the squares can meet in two states: black or white, which in the context of the game mean dead or alive. There are simple rules for the emergence and the disappearance of the elements according to the number of “neighbors” of each element. The observed behaviors consist of elements which, in group, exhibit certain configurations at first not programmed, but which derived from the rules of the program.

The theoretical research pointed out scientific efforts in search for algorithms that worked in an organic manner, and that were more complex and closer to real forms of organization, including live forms. It also pointed out how the relationship between artistic creation and technique can happen in different critical ways that show the difference between creation in art and creation with strictly pragmatic objectives.

Programming languages consist of a text that has the distinction of being feasible. There is a long cultural history of the feasible text even before the computer. From Raimundo Lulio, to Oulipo, and even to the most recent uses of code and programming aiming at the creation of experiences, there is an imagination that serves to visualize powers of creation for the language between their structural features and their perceived sense. Lulio (Catalan philosopher and missionary, 1233-1316) initiated a longstanding tradition in search for the generation of meaning from formal operations. His intention was to create a system that generated valid premises for all cultures and religions. This system was created in the form of a combinatorial table, but what could not be avoided was the generation of false premises that were as predicted as the true ones. After Lulio, many philosophers and linguists attempted to create “machines” that generated semantic structures to generate, for example, all the words of a specific language or combinatorial poems in which few verses would generate a large volume of writing. However, the efficiency of these machines has always been questionable because eventually it reveals the mismatch between the strictly formal and the semantic, as Lulio discovered in his experience.

Oulipo (Ouvroir de Literature Potenciel), group which included Raymond Queneau, Italo Calvino and Georges Perec, emerged with the idea of relating the creation through language with formal mathematical rules. Oulipo sought to make creative works in spite of literary constraints. The rules served for the creation with them and in spite of them. For this group, whereas the rules favored creation, they also offered obstacles to it. Still, rules are their starting point and work as the self-imposition of an algorithm for creation.

Among the proposals that use computational interfaces, their codes and programs, the one that stands out is the critical use that the Jodi duo (Joan Heemskerk and Dirk Paesman) makes of the web and its formatting language, HTML. In their works, they do not highlight the efficiency of these means but the errors that can originate from them. The duo’s website is frequently perceived as a dangerous area of the web: a website full of viruses; a curious web address (http://wwwwwwwww.jodi.org/); and a chaotic mass of green flashing characters that is presented in the home page. Proposals such as this one question the inevitability of the error, even in an environment in which control and organization have a fundamental role. We perceive Jodi’s work as an error because we have already experienced the errors of the machine. In proposals of web art, what is frequently highlighted is not good navigation, but the labyrinth into which the intricate web of the internet links transforms, in our perception. The structure of the web favors the intersection of these pieces of information from so many diverse natures and the proposals of web art play on this feature.

In the examples above, we saw ways in which imagination acts upon and with language, as well as ways through which we perceive these actions. In this proposal there is the intersection of formal rules of programming, the two-dimensional language of drawing and its planning in terms of automation. This proposal was initially thought as web art, with a hypertext structure and links between images that work as interfaces and would give the idea of a route to be followed by the programming languages and their potentialities in the organization of images. From there, the research focused on these potentialities. The destination of the programs developed in this research will still be the internet, only that now they will be in the form of programmed drawings, only broadcast online.

The research about formal potentialities in the composition of drawings continues with the study of the algorithms’ possibilities and in the attempt to generate drawings out of chaotic but consistent appearances and behaviors. In the process of writing programs that compose designs, several structures and behaviors still need to be tested. The challenges include the creation of structured compositions that show apparently unstructured images, programming design-interfaces that work in apparently unexpected ways, and confronting formal structures to apparently chaotic designs. In sum, the challenge is to organize complex designs in which perceived order and disorder coexist and to work in the limit between rules and sensitivity.

NOTE

1. In Portuguese there are no words to distinguish design from drawing.

BIBLIOGRAPHY

ABOUT THE AUTHOR

Angélica Beatriz