1. Marguerite Yourcenar, The Abyss (New York: Farrar, Straus, and Giroux, 1981), p. 192.
2. "From this day, painting is dead" is what the French painter Paul Delaroche reportedly said in 1839, when confronted with photography for the first time. See Nancy Roth, "Art's New Address," in The Techno/Logical Imagination: Machines in the Garden of Art (Minneapolis: Intermedia Arts Minnesota, 1989), p. 6.
3. Paintings and drawings reproduce only the essential elements determined by the artist. In fact, the drawing process by which an image is created is an act of constituting the significant from the insignificant. Rodin argued that: "It is the artist who is truthful and the camera that lies because a photograph is a mechanical representation frozen in time. The photograph is false because in reality time does not stop. If the artist can succeed in representing the fluidity of motion in a painting, the work is much more original than the scientific [photographic] image in which time is abruptly suspended." Paul Virilio, La machine de vision (Paris: Editions Galilée, 1988), pp. 14-15.
4. Irene J. Winter, "After the Battle Is Over: The Stele of the Vultures and the Beginning of Historical Narrative in the Art of the Ancient Near East," Studies in the History of Art, vol. 16 (Washington: National Gallery of Art, 1985), p. 28.
5. Norman Bryson, "Vision and Painting" (New Haven: Yale University Press, 1983), p. 14.
6. Umberto Eco, "How Culture Conditions the Colours We See," in On Signs (Baltimore: John Hopkins University Press, 1985), p. 163.
7. Bryson, op. cit., p. 13.
8. Linda Hutcheon, "The Politics of Postmodernism "(London: Routledge, 1989), p. 123.
9. Michael Shapiro, "The Political Rhetoric of Photography", delivered at the 82nd Annual Meeting of the American Political Science Association, Washington, D.C., August 2831, 1986.
10. Bryson, op. cit., p. 14.
11. Edmond Couchot, "La synthèse numérique de l'image," Traverses/26 (Paris: Centre Georges-Pompidou, 1983), p. 60.
12. From an interview published in "The Work of Art in the Electronic Age," Block 14 (London), autumn 1988, p. 5. Virilio gives as an example the works of the mathematician Netgrin, whose computer programs pictured Moebius rings turning.
13. Couchot, op. cit., pp. 58-62
14. Howard Bossen, "Zone 5, Photojournalism, Ethics, and the Electronic Age," Studies in Visual Communication (University of Pennsylvania) 11:3 (summer 1985).
15. "When an artist uses a conceptual form of art it means that all the planning and decisions are made before and the execution is a perfunctory affair . . . This kind of art is not theoretical or illustrative of theories; it is intuitive, it is involved with all types of mental processes." Sol Lewitt, L'art conceptuel: Une perspective (Paris: Musée d'art moderne de la Ville de Paris, 1989), p. 199.
16. Roland Barthes, "The Rhetoric of the Image." in Image-Music-Text (Glasgow: Fontana Collins, 1977), p. 39.
17. Theodore Roszak, "The Cult of Information" (New York: Pantheon Books, 1986), p. 16.
18. The equation, which I had found in an article titled "From Noise Comes Beauty," by Carl A. Pickover, a scientist at the IBM Thomas J. Watson Research Center in New York, was published in the March 1988 issue of "Computer Graphics World".
19. John F. Asmus, "Digital Image Processing in Art Conservation," BYTE 12:3 (1987), p. 151.
20. On September 18, 1988, in Budapest, I videotaped a public demonstration of about 80,000 marchers who were protesting the ecologically unsound Nagymaros/Danube hydroelectric works. The project was engineered by the Soviet Union, which contracted with Austrian firms for its construction on Hungarian and Czechoslovak territories. Because of the participation of the Austrian Green Party, this was the first demonstration to be tolerated by the officials in Socialist Hungary.
21. Claude Lévi-Strauss, "The Savage Mind" (Chicago: University of Chicago, 1966), p. 22.
Image, Language, and Belief in Synthesis
The sixteenth-century alchemist Zeno, catching himself in the act of seeing, was shocked by the unusual sight of his own eye mirrored by a technology: a magnifying glass that he used to examine the plants he collected. Zeno's startling experience parallels the complex relationship of technology and human consciousness technology as an extension of the human body, as a mirror of the self, as a mediation between nature and culture, as a potential discursive medium or a tool of alienation and control.
All technologies distort. By expanding our abilities to perceive, they simultaneously diminish us. We experience the world through the senses and the act of seeing is one of giving meaning, taking stock of our environment to counterbalance chaos. Technologies that help us to see shape the way we see, and, in the end, determine how we see. These inventions have resulted from choices framed by cultural beliefs to arrive at a particular view of the world, not representing the totality of human experience but a view locked within the limits of a fluctuating history. In the way that we are born into language, we also enter an unfolding, socially defined world of visual continuum. We integrate these conventions unquestioningly, recycling them in varying degrees as a means to arrive at the new. As we consider the impact of digital technology on the production and interpretation of images, questions arise about the belief systems that are in place and their development over time.
Images through Western history have functioned to convey beliefs, becoming authoritative records by making permanent the transitory. With the introduction of high-resolution still and motion photographic representation, images have maintained their status as the dominant mode of information exchange where visual documentation is of major importance as for instance in television news. Contemporary theoretical discourse has dealt extensively with the subjectivity inherent in photographic representation. It is now generally accepted that, even though the photograph represents everything in front of the camera, photography is a symbolic practice where meaning is determined by beliefs and generated through the connotative strategies of subject selection, framing, and vantage point.
The polemic between photography and painting in the 1860s may be a useful reference point regarding the impact of digital processing on the interpretation of images. Photography's mechanical mode of optically recording reflected light onto a light-sensitive surface was initially accepted as a freedom from intervention.2 One could visually perceive the resulting differences between the literal mechanical recording of the camera in relation to the coded (stylized) painted work of the same scene.3 As a result, it exposed the inherent subjectivity in painting and forced painting to recognize its function as a distillation of an experience in perceiving instead of a depiction of a subject for itself. Digital technology hinders this type of debate as long as the criterion for verisimilitude is based on appearances since digital processes can simulate existing conventional media to the extent of being visually indistinguishable from them. In the process of converting analog data, such as a continuous-tone image, into digital form, a fundamental transformation takes place. Once the image is stored as numeric data in computer memory, it can be processed in unlimited ways without degradation of information or any trace of change. Given the very high probability that digital filtering of one sort or another could have been used in the transmission process, prior knowledge about an image's history, its source, mode of production and reproduction have become necessary informational components to accurately understand the full meaning of a digital image.
The media through which cultural images are processed and transmitted are influential components of the visual narratives conveyed. The technological tools of production could not exist outside of institutionalized, ideological constructs since their inventions and utility are socially determined. In digital processing, as in any other form of communication, the technological components of hardware and software are structures that shape and impose a form onto the information they process, but these mediating structures are normally understood as transparent or "value free." In an evaluation of the function of narrative in iconography, the historian Irene J. Winter states that "one must divide the message into at least two components: the actual information conveyed, and the extralinguistic or extravisual referent that is part of the subtext. The ideological message is often built into the structure of how the message is conveyed, rather than what the message contains."4 In the everyday usage of a tool, as with the everyday acceptance of images and language, little thought is given to their particular, ideologically determining, functions. In fact, their success in naturalizing the beliefs of a given community depends on the degree to which they remain unknown as independent forms.5
The act of seeing in everyday life is generally taken as a natural event. We trust what we see without much questioning as to why we give importance to certain things and why we sometimes do not recognize visual evidence presented in front of us. According to Umberto Eco, one can usually communicate only about those cultural units that a given signification system has made pertinent.6 Moreover, competence in reading visual imagery is an acquired skill similar to the process of learning language, a social activity defined by the norms of a particular culture. Norman Bryson maintains that the reality experienced by human beings is always historically produced. That it is "more accurate to say that realism lies in a coincidence between a representation and that which a particular society proposes and assumes as its reality; a reality involving the complex formation of codes of behaviour, law, psychology, social manners, dress, gesture, posture all those practical norms which govern the stance of human beings toward their particular historical environment."7
Photography, Digital Processing and Social Practice
In our image-saturated society, our cultural myths and beliefs are daily reinforced through the numerous photographs we come across from advertisements, to the news, to family-vacation snapshots. Photography's apparent transparency promotes a viewing experience that arouses pleasure without creating any awareness of its act of ideological constructing.8 It is easily accepted as a window on the world rather than as a highly selective filter, placed there by a specific hand and mind. The photograph is treated so unproblematically as "real" that its grammar of discussions tends to approximate the grammar of face-to-face encounters:9 "This is Peter Jennings" is a culturally accepted statement whether one is presenting the person, or pointing at a shape on a television screen.
Photography depends on the physical world as raw referent resource, inasmuch as it requires a subject in front of the camera. Its meaning is culturally defined. Practice and beliefs dictate the image maker's decisions about subject selection, framing, and the moment(s) of exposure. The viewer's understanding and responses are defined in turn by a process involving implicit cultural knowledge that exists nowhere in codified form but remains at a tacit level.10 As a result, optically recorded visual communication becomes a highly effective rhetorical tool, as it tends to have a persuasive impact. Photographic images imply the potential for verification. There is the general presumption that the image must have been dependent to some extent on a real-world event. Concrete information creates belief.
Optically recorded images such as photographs and video imagery consist of minute, indeterminately arranged components such as the chemically generated grain in film and the variation of light intensities in video. While the digital photograph looks like its conventional counterpart, when examined very closely, it reveals itself to be composed of discrete elements called pixels, which are assigned precise numerical values. Each pixel in the image has a determined Cartesian horizontal and vertical location value and a specific colour-intensity value. It is this relationship of modular units with definite values that makes it totally controllable. Thus, when we speak of the digital photographic image, we are referring to a simulated photographic representation, achieved through any combination of a mechanical lens, a hand-held (electronic) pencil, or a database filtered by mathematical language.
A digital, numeric-based structure is by definition a statistical representation where the degree of accuracy is dependent on the amount of information that can be processed within a given space and time. The greater the memory, the richer the image's degree of resolution. When the volume of statistical data surpasses the threshold of our physiological capabilities to perceive change, the illusion of total simulation is achieved. A digital image does not represent an optical trace such as a photograph but provides a logical model of a visual experience. In other words, it describes not the phenomenon of perception but rather the physical laws that govern it, manifesting a sequence of numbers stored in computer memory. Its structure is one of language: logical procedures or algorithms through which data is orchestrated into visual form. Even though both may look the same on the surface, a digital image may be said to differ from its analog counterpart in terms of a verifiable past and a possible future. Because of its dependence on an a priori, real-world referent subject, a photograph by nature refers to the past a viewing experience termed by Roland Barthes a sense of the "having-been-there." With the digital image, whose construction could potentially be totally fictive, one can claim at most that the event represented "could possibly be."11
The Medium of Mathematical Language
Digital images simulate rather then represent the real. They can be produced from total invention according to mathematical algorithms, making visible, through the use of computers, concepts and physical phenomena that do not exist in material form. For instance, objects and images that are rotated on television broadcasts exist in virtual space an environment that is totally fictional, defined mathematically, and based on laws of physics. Paul Virilio describes the digital, synthetic image as a tool for seeing the invisible, things you cannot see in any other manner than by calculation.12 This displacement of the real by simulation shifts the image's status to a total representation of concepts. One first conceptualizes one's intentions, then proceeds to realize the imagined through programming.
Computers, which are symbol-manipulating machines, effect a radical rupture with conventional approaches to image-making. For artists who create through programming, logical language mediates between intention and the resultant artwork by a process similar to the ways composers create through musical notation. The working method is divorced from sensory experience in that the artist's work becomes one of orchestrating symbolic order through code writing rather then through the physical interaction between material (such as paint) and the senses.13 It forces the artist to conceptually translate events from the real world into complex sequences of rule-based decisions and to previsualize in so precise a manner that even chance needs to be determined and coded. Once these logical sequences and commands are stored in memory, any aesthetic and/or logical errors can be reformulated by simply changing the code, because computers provide the ability to retrace and undo one's steps.
Some Philosophical Questions
The processing of digital information depends not on generational reproduction, which would imply data loss, but on the transference of a sequenced set of numbers by which information can be duplicated ad infinitum. Since each individual element of a digital image is readily accessible, the copying capability has forced a reevaluation of what constitutes ownership and authorship. For instance, is a change of one pixel in a high-resolution digital image a significant alteration to authorship? In this age of information exchange, where a shift in context results in a shift in meaning, the issue would seem to be one of intention and positioning. Sociologists have voiced concerns about the potential misuse of digital technology in the news media pointing out that "those who have access to digital image-processing systems have the capacity to alter, reconstruct, or create imagery reflective of the real world that might be passed off as representing accurate data."14
While photographs have from the start been doctored or their meaning changed through standard photographic strategies of vantage point, framing, cropping, and retouching, the digital image betrays no surface evidence of alteration. One must first suspect that the image is less than accurate, then one needs a computer with the right program to detect the changes.
The focus of my artistic work in photography during the past fourteen years can be summarized as a series of investigations about the cultural and syntactical conventions by which photographic images are structured to convey meaning. Under the Spreading Chestnut Tree (1984), examines the coding conventions of corporate iconography. In this work, the compositions and body language of group poses in the photographic portraits of the 1980 E.F. Hutton annual report were reconstructed in the tradition of the "nature morte" or still life. The figures were executed to look like wood veneer, a material commonly associated with the corporate body through its use in the architectural and design finishes of corporate boardrooms. These constructions were then photographed by traditional means and printed on metallic gold and silver paper. I began working with computer programming in 1981, realizing at that time that aesthetic decision-making procedures could be formulated in terms of a sequence of logical conditional statements not unlike rules of language. This structured approach has had precedents in the sequential and modular artworks of Sol Lewitt, and the syllogistic propositions of Lawrence Weiner, Doug Huebler, Joseph Kosuth, and others associated with the conceptual art movement.15 The fact that the construction and manipulation of photographic images could take place visually on the computer screen rather then materially in front of the camera lens has made computer work an effective alternative.
News Beirut, 1987 from the Words & Words Series considers the discursive function of textual labelling in TV news imagery. Frames from the news were digitized and segments that contained text were isolated and used to "paint in" the whole image area in a random fashion. These labelling texts names of cities where the particular news event took place function in television broadcasts to contextualize images. They impose a meaning that may not necessarily pertain to the original intent of the image. Barthes calls this a strategy of anchorage: "The caption helps me to choose the correct level of perception . . . , focussing my understanding. The anchorage is ideological. It remote controls the viewer toward a meaning chosen in advance."16 By erasing and covering the image with fragments of the text, the text becomes the aesthetic experience, still maintaining a connotative load. As in the original, we project onto the new image a preconceived idea of what that text refers to. Our perceptions about the world we live in are reinforced by the conventions of representation that inundate us daily, and television's particular mode of defining reality seems to be a dominant conditioning force within the culture. Television becomes an unlimited, real-time, image bank when it is linked to an image-processing computer. Appropriating images from this data bank is a means to examine and comment on television's highly ritualized syntax.
A conceptual starting point for investigating the potential of image-processing software that I have been producing since 1986 is Claude Shannon's information theory, which has been described as "mathematics turned into philosophy." My approach has been to reverse that sequence, commenting on the semantic discrepancies of information theory, specifically in its definitions of language, signal, and noise. Noise in information theory is defined as random errors within a signal or unstructured information, distinguishable from signal that is ordered information. Whereas engineers have invested much energy in filtering out noise from signals to purify communication, my programming activities aim to achieve the opposite to incorporate into images an order determined through noise and chance. The program Xerox replaces the "inessential" components of an image with a noise pattern, exemplifying one of information theory's dictates that there is greater than fifty percent redundancy in language in the form of sounds or letters that are not strictly necessary to convey a message. The resultant image looks as if it has been photocopied, with over fifty percent of the pixels reduced to black. The actual percentage of "value-free" information (eighty-five percent black in the work illustrated) is calculated at the end of the process.
In Moral Stories, a program titled Smudge removes image sharpness and photographic depth by a random process of blurring image areas over time. The end result, which resembles water spilled on an ink drawing, looks as if it were hand painted. When the process is prolonged, it results in the annihilation of the image in a total blur. Although information theory has revolutionized our culture by demonstrating that information is a quantifiable entity which can be calculated and controlled, thereby providing the theoretical basis for the development of new technologies and modes of information processing, it has also semantically shortchanged our understanding of the value of things communicated. In the bipolar reduction to either signal or noise, the meaning of things from moral injunctions to philosophical treatises, from love poems to sales messages and parking tickets is levelled into signal. So, too, their value. Theodore Roszak comments: "Thanks to the high success of information theory, we live in a time of blinding speed; but what people have to say to one another by way of technology shows no comparable development."17
A recent project titled Between East & West (1990), brings together the logistics of software development within a context that is both personal and political. It involves a working dialogue with a scientist from Hungary, where I was born (and which I left in 1956 during a time of crisis). The project began with an exchange of information: I sent my colleague a mathematical equation to be turned into computer code, a program that would produce images.18
After generating a series of random numbers, the equation assigns the average of a "neighbourhood" (a cluster of visually adjacent units) of values to each pixel, thereby arriving at a balance between chaos and order. This algorithmic process, derived from image-processing filters, is used to sharpen photographic images. It has particular application in specialized fields where photographic recording provides data documentation, such as the space program, surveillance monitoring, and recently art conservation.19 Similar digital filters have also been employed to enhance features of the Shroud of Turin, sometimes described as the "first photographic phenomenon in history." For some time, I had been aware that dialogue in various professional fields had been taking place between the West and the East prior to the recent political changes. That scientists and specialists were communicating on a regular basis, sharing their knowledge and exchanging information. This reality seemed to be in sharp contrast with the official political positions of both sides.
My intention in re-enacting this informational exchange has been to make visible the unspoken relationships and to bring attention to the influence of Central European intellectual ideas in Western thought, something that has largely been ignored within the culture at large. I retrieved the program on a visit to Budapest in August 1989 and expanded it in California to implement colour values. These image-enhancement techniques are usually applied by first digitizing images, then filtering them with the algorithmic functions to remove noise and highlight detail. In East & West I used the functions instead to generate abstract noise patterns directly from the computer program without employing a referent photograph in the process. In the center of each textured image was placed a long rectangular panel that looks like a metallic nameplate. Constructed in computer memory through software, it was programmed to have the reflecting properties of chrome, on which images from a video I made in Budapest were "projected" mathematically.20 The viewer would not be able to decipher what the plates reflected, but because of our general familiarity with the photographic image, the blurred shapes did read as photographic and therefore maintained the authority associated with photographic representation.
The impetus for this work came out of considering the belief structures we bring to the reading of images. Works of art, like other signs, are initially empty but gain their meaning within a historical context, mediated by current social belief systems. In the case of the East & West project, where the only existing element in the image is the pattern created by the filter (minus the photograph), the works become similar to the skeletal remains of a stream of punctuation marks after words have been emptied out of a sentence.
In his description of "bricolage," Claude Lévi-Strauss makes reference to the artist as someone who constructs through craftsmanship a material object, a symbolic structure that is also an object of knowledge.21 The act of craftsmanship in my work of the last few years has been situated in the programming phase, i.e., the writing of computer code, this being the arena where the relationship of language to image and the impact of technology on cultural vision is dealt with. The weight of this activity remains largely invisible, resulting in images that function as outward manifestations of the locus of the work: the software itself. These images serve as residual physical indices to an event that has taken place.
With computer technology, a new twist is introduced to our acceptance of visual evidence: photographic representation can be fabricated through mathematical processing. Computer imaging can simulate and generate any type of imagery or "reality" one chooses. The ideological factor in the digital photograph is exponentionally increased since the image promotes a form of meaning and value while appearing to merely represent the real. In my computer-generated works, as in my investigations of photography, the intention is to engage the viewer in considering the discrepancy inherent in an image that looks natural on the surface but is in fact mediated, and is therefore a challenge to conventional notions of belief in visual representation.