About the Author
Read an Excerpt
A History of Vision and Reason since 1945
By Orit Halpern
Duke University PressCopyright © 2014 Duke University Press
All rights reserved.
Temporality, Storage, and Interactivity in Cybernetics
Few figures are more prominent in the histories of digital media then the MIT-based mathematician Norbert Wiener. From 1950 until the late 1970s, his work was prominently featured across multiple fields from architecture to sociobiology. For almost thirty years after World War II, before the term "digital" gained prominence, cybernetics, a word he coined, was the language used to describe a transformation in life and a new technical condition related to, but not reducible to, digital computers. In the late 1990s with the advent of the Internet, his name returned in the effort to historically situate the rise of digital networks and the interactive interface.
Wiener almost appeared to anticipate his future popularity. His archive at MIT is a fascinating exemplar of a life turned into data. Carefully curated, every letter mimeographed and saved, it is as though Wiener was already preparing his life for transmission, assuming a seamless translation between personal experience and historical analysis. Fastidiously cataloguing his many failures in natural history and the sciences of empiricism and experiment, he turned to reformulating these experiences in service of another form of knowledge. Rather than speak of the value of personal experience or the specificity of his character, he sought to make that element of his innermost psychology—his character—the substrate for legitimating computation.
I wish to take up this turn away from an "external" world and the devolution inward, in this case to the very self, as a starting point to consider the relationship between the archive and the interface in digital systems. What might we make of this move from a concern with recording an external, perhaps "natural," world in its entirety to an obsession with processing the already recorded traces of memory? How do we wish to frame this shift to forms of representation whose reference is reflexive rather than indexical? Wiener was not naïvely recounting his failures in finding adventure, his inability to excel in the life sciences; rather he was articulating an aspiration for forms of technology—both of thought and machine, or perhaps of thought as a machine—that had not yet come into being when he spoke. In his work, and that of his many compatriots in the arts and sciences of the time, we hear similar statements that voiced a not-yet-realized aspiration to transform a world of ontology, description, and materiality to one of communication, prediction, and virtuality. A world that, perhaps, speaks to our contemporary fantasies of a data-filled space where every screen is an interface, every diagram a process.
But if Wiener attempted to propagate the "new," it only came into being through the memory traces of the old. It was by way of Freud, the exemplar of a previous century's sciences, that Wiener implied the impossibility of describing a world in its totality, of ever rendering "reality" legible. Instead, he argued, we are faced with an "incomplete determinism," an operative lack that cannot enter description but can produce something else—a self-referential and probabilistic form of thought:
one interesting change that has taken place is that in a probabilistic world we no longer deal with quantities and statements which concern a specific, real universe as a whole but ask instead questions which may find their answers in a large number of similar universes.... This recognition of an element of incomplete determinism, almost an irrationality in the world, is in a certain way parallel to Freud's admission of a deep irrational component in human conduct and thought.
This form of probabilistic thought that emerged at the turn of the last century would now, in Wiener's work and that of his compatriots in the information sciences, be connected with theories of messages. Wiener was comfortable with acceding that the universe in its plurality might never be known. This accession, however, was only made to allow for the possibility that within far more localized situations, the future—chance—might yet be contained by way of technology.
But Wiener's invocation of Freud also complicated his own vision for technology and science. His statements posed the possibility that the contemporary systems he hoped to bring into being were not absolutely amnesic to their history. His statements would be, and still are, haunted by the residual problems of recording, translating, and transmitting information and associated concerns with indexicality, signification, and representation. Unconsciously, perhaps, even Wiener acceded to the possibility that not all forms of information could be similarly recorded and transmitted without loss, transformation, or change. It is precisely at this site, where the traces of older histories mark the desire for the production of the new, that I will excavate in this chapter.
Wiener's texts, and the work of his compatriots in cybernetics and the neurosciences, serve as useful vehicles, therefore, to begin investigating this historic attachment and displacement of older technical questions of documentation, inscription, and perception into terms of information and communication. The relationship, explicitly detailed in the work of many early cyberneticians, between the record, the diagram, and communication forms a bridge between our contemporary discourses about archiving, screens, and interactivity and historical concerns with memory, temporality, and representation. At this pivotal moment, demarcated by a catastrophic world war, these sciences were part of producing an aspiration for a new world constituted of information; but not without producing a novel set of conflicts, desires, and problems. I turn, then, to outlining what the conflicted relations between the archive and the screen might still have to say to our desire for "interaction" and communication with and through our machines.
Cybernetics: Communication and Control
The very definition of cybernetics already assumes a complex relationship to temporality and history—bridging the past with an obsessive interest in prediction, the future, and the virtual. As the etymology of the word suggests, cybernetics is a science of control or prediction of future action. In further adjoining control with communication, it is an endeavor that hopes to tame these futures events through the sending of messages.
These rather abstract ideas of communication as the source of control consolidated themselves within the milieu of military research and development in antiaircraft defense systems during World War II. While scientific research has long been part of war, World War II is now widely heralded as marking a critical turning point in the organization of science both in scale (billions rather than millions allocated) and in the wholesale recruitment of both industry and academic researchers. To offer some substantive examples, major beneficiaries of the effort included MIT (an institution already active in recruiting military spending in the interwar period), which was awarded some $117 million in R & D contracts (approximately 1.6 billion in contemporary dollars) and remained among the top sixty defense contractors after the war); Caltech, with $83 million; and Harvard and Columbia, with about $30 million each; compared to $17 million for Western Electric (AT&T), $8 million for GE, and less than $6 million each for RCA, DuPont, and Westinghouse. Many other industrial groups also received assistance, including Bell Labs, IBM, and Norton. In sum total the U.S. Office of Scientific Research and Development spent $450 million (approximately $5.5 billion in inflation-adjusted contemporary dollars, although probably far more if other indexes of purchasing power or dollar value are used).
Part of this reorganization toward what is now labeled "big science" was a transformation in disciplinary boundaries. Wiener worked in one corner of what was quite literally a very large complex. The Radiation Laboratory (commonly called RAD lab) at MIT was among the biggest of the American research installations. The space was costly at $1 million, and contained cutting-edge equipment in radar signaling, calculation, and communication technology. Housing at its pinnacle over four thousand different researchers from around the world, the site was host to engineers, physicists, mathematicians, physiologists, doctors, and individuals from a range of other fields, all investigating, in one way or another, signal processing and communication.
Built barracks style, out of wood, a haphazard block of extensions and convoluted hallways, it was, to quote one of its inhabitants, the future neuroscience and perception researcher Jerome Lettvin, "the womb of ideas. It is sort of what you might call the vagina of the institute. It doesn't smell very good, it is kind of messy, but by god it is procreative, and it doesn't make only replicas of itself, as other buildings do. It is sort of all-purpose." As Stewart Brand, of Whole Earth Catalogue and Whole Earth 'Lectric Link (WELL) fame, writes, RAD lab was a sort of emergent architecture, a haphazard and growing set of wooden boxes and storage crates that simply grew and adapted, formlessly, to the different interactions of its inhabitants. Inhabitants reminisced about the way engagements would simply occur by accident and haphazardly in the hallways between the shed-like segments of the lab.
But for all the purported camaraderie and creativity attributed nostalgically to the lab, the milieu was a high-stress space where the topic at hand was violence at a distance, particularly that coming from aerial warfare with its velocity and territorial reorganization. Within this context and under the imperative of rapid defense, Wiener, working with neurophysiologists and doctors and influenced by early work on computational machines (the differential analyzer), argued that human behavior could be mathematically modeled and predicted, particularly under stress; thereby articulating a new belief that both machines and humans could speak the same language of mathematics.
By reformulating the problem of shooting down planes in the terms of communication—between an airplane pilot and the antiaircraft gun—Wiener and his compatriots hoped to devise better defense systems. Wiener, working with the MIT-trained electrical engineer Julian Bigelow and the physiologist Arturo Rosenblueth, under the guidance of the director of the Applied Mathematics Panel at the Office of Scientific Research and Development, Warren Weaver, decided to treat the pilot of a plane as a machine. These researchers postulated that under stress airplane pilots would act repetitively, and therefore have algorithmic behaviors amenable to mathematical modeling and analysis. To this end, they designed experiments to simulate flying through the turbulence, noise, and antiaircraft fire of a bombing run. Wiener and his colleagues set up a fake cockpit-type scenario, where a "pilot" was asked to guide a light beam through a particular set of invented obstacles. They then gathered data about the movements the individual made. They discovered that while different "pilots" appeared to have widely different patterns, for any one individual, there was a striking amount of redundancy and repetition in the movement. People do, it would appear, act quite mechanically under duress.
Behavior, Purpose, Teleology
For a project so literally obsessed with problems of response time and the description of enemies, it is of little surprise that ontology (or the identification and definition of the enemy), representation, and temporality should be explicit and central concerns. What is somewhat less intuitive is how these questions of immediacy and identification became ones of storage and communication.
There was a series of moves by which this new world of perfect communication between networked entities emerged. The early efforts to rethink technological defense mandated rethinking what an enemy was, rethinking what communication was, and isolating the communicative relations between enemies in a manner that allowed them to be modeled and abstracted into computational methods. This wartime research engine therefore produced what the historian of science Peter Galison has defined as a new "ontology of the enemy," not the alien and animal opponent, not the distantiated space on the map of an airtime raid, but the "cold-blooded, machine-like opponent." This move eliminated the enemy as visibly different and produced an imagined closed world of networked communications between informatic entities. This emergent assemblage would be codified later under the Cold War ideal of "C3I: command, control, communication, and information." In this new imaginary, response time was the critical feature driving the system, and this produced a new understanding of teleology that was not about progress, linearity, or conscious humanist effort, but rather a new mode of technical thought. In fact, it is arguable that Galison's use of the term "ontology" is misplaced. Cyberneticians concerned themselves with process, not essence. Within this rubric, visually representing the enemy was not the core concern, but rather the training of the machine or person to communicate and anticipate further signals.
The first major concept was to introduce the idea of black boxes. For the antiaircraft project, engineers and mathematicians, using early computing devices, began to view the gun and the plane behaviorally—which is to say that while the internal organization was opaque, unseen, and potentially different, the behavior or action was intelligible and predictable. The world, therefore, became one of black-boxed entities whose behavior or signals were intelligible to each other, but whose internal function or structure was opaque, and not of interest. We can say that this view of the world concentrated on process, not structure or difference. Julian Bigelow, who worked with Wiener to produce and articulate many of these ideas, put it succinctly when describing his fellow researcher: "Wiener thought as a physicist really. He thought in terms of process. He thought in terms of some kinds of physical models or some kinds of intuitive models." These personal descriptions fit the larger project, as well, where precision was not valued, but rather the development of technologies that could be extracted, moved between locations, and do work. Wiener would argue, in the terms of gestalt, that the questions of perception, for example, were simply the abstraction of forms from the world, and that those forms were not unique to a particular sensory operation, "We have thus designated several actual or possible stages of the diagramatization of our visual impressions. We center our images around the focus of attention and reduce them more or less to outlines. We have now to compute them with one another, or at any rate with a standard impression stored in memory.... This is a general principle, not confined in its application to any particular sense and doubtless of much importance in the comparison of our more complicated experiences." Behind the black box is the concept of a world that can be remade into exportable processes that point away from their initial sites of inception.
The second idea introduced by Wiener and his colleagues in their wartime research reworked the idea of communication and interaction into terms of statistical prediction and feedback. The capacity to predict a black box's action came to be known as feedback. Wiener would define feedback as "the property of being able to adjust future conduct by past performance." It was argued that all behavior culminating in a goal was purposeful, teleological, and the result of negative feedback, or the need to correct the input signal. For example, a missile seeking a target swings afar and misses then adjusts its flight path in response to the signal given from its target. Negative feedback assumes that both entities are in relationship to each other, are both communicating, and are both changing their behaviors in relation to the other.
To appreciate what such an understanding of feedback implies mandates that we ask what, precisely, is being redefined in the ideas of "purpose" and "teleology." Purpose was now understood as any active behavior that was not random but rather "meant to denote that the act or behavior may be interpreted as directed to the attainment of a goal—i.e., to a final condition in which the behaving object reaches a definite correlation in time or in space with respect to another object or event." This was quite explicitly delineated from, say, a watch, which has a given task—keeping the time—but no set or defined endpoint to this function. Teleology was a concept simply adjoined to any purposeful behavior of entities engaged in negative feedback (or dually responsive) interactions.
The result of this logic was that teleology was considered no longer "causal," and purpose no longer conscious, in that there is no given chain of events that consciously leads to a predetermined goal. In the seminal 1943 article consolidating these ideas—"Behavior, Purpose, and Teleology"—Wiener, Rosenblueth, and Bigelow argued that "the concept of teleology shares only one thing with the concept of causality: a time axis. But causality implies a one-way, relatively irreversible functional relationship, whereas teleology is concerned with behavior, not with functional relationships." Which is to say that while both types of interactions occur within time, and are irreversible, one is simply assessed by any interaction that produces an effect, and the other—the functional and causal—is deterministic, it has only one possible outcome. In a functional interaction, A must cause B and only B to occur, whereas in this new "purposeful" and behaviorist approach there is a cause and effect, but the relationships between cause and effect are not predetermined; rather, they are a choice, a likelihood, the result of a series of possible interactions. Change, which constitutes teleology in this system, occurs through a series of repeated and predictive actions. This is a space of encounter between two like objects communicating and responding to each other, forming a closed world of predictive encounters.
Excerpted from Beautiful Data by Orit Halpern. Copyright © 2014 Duke University Press. Excerpted by permission of Duke University Press.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
Table of ContentsAcknowledgments vii
Prologue. Speculating on Sense 1
Introduction. Dreams for Our Perceptual Present 9
1. Archiving. Temporality, Storage, and Interactivity in Cybernetics 39
2. Visualizing. Design, Communicative Objectivity, and the Interface 79
3. Rationalizing. Cognition, Time, and Logic in the Social and Behavioral Sciences 145
4. Governing. Designing Information and Reconfiguring Population circa 1959 199
What People are Saying About This
"Beautiful Data is a wonderful book, deeply engaging and full of compelling insights. Reading across fields, disciplines, borders, and issues, Orit Halpern chronicles the emergence of a new way of thinking about the world for the digital moment. It is crucial reading for anyone interested in the new directions in which the humanities, the arts, and education are moving."
"From the title to the last page, Orit Halpern experiments with a heady mix of memory, speculation, and the physical world. Beautiful Data starts with the early days of cybernetics, back when the nascent discipline was undisciplined, roaming through the world, as much about architecture and design as it was about mathematics, physics and the functioning of the brain and body. Halpern then pushes on that openness, exploring design in the work of Kepes and Corbusier, up on through the vast new Korean smart-city Songdo, always returning to the control of data as it restructures our archival past and sketches our possible futures. An ambitious book, Beautiful Data is like a light pipe, pumping an ever-changing flow of new ideas about data and feedback in sudden and productive combination."