Liberal Subjectivity Imperiled: Norbert Wiener and Cybernetic Anxiety

Of all the implications first-wave cybernetics conveyed, perhaps none was more disturbing and potentially revolutionary than the idea that the boundaries of the human subject are constructed rather than given. Conceptualizing control, communication and information as an integrated system, cybernetics radically changed how boundaries were conceived. Gregory Bateson brought the point home when he puzzled his graduate students with a question koan-like in its simplicity: "Is a blind man's cane part of him?" [1] [2] The question aimed to spark a mind-shift. Most of his students thought human boundaries are naturally defined by epidermal surfaces. Seen from the cybernetic perspective coalescing into awareness from during and after World War II, however, cybernetic systems are constituted by flows of information. In this viewpoint cane and man join in a single system, for the cane funnels to the man essential information about his environment. Similarly for a deaf person's hearing aid, a voice synthesizer for someone with impaired speech, and a helmet with a voice-activated firing control for a fighter pilot.

This list is meant to be seductive, for over the space of a comma, it moves from modifications intended to compensate for deficiencies to interventions designed to enhance normal functioning. Once this splice is passed, it becomes difficult to establish conceptual limits to the process. In "A Manifesto for Cyborgs," Donna Haraway writes about the potential of the cyborg to disrupt traditional categories. [3] Fusing cybernetic device and biological organism, the cyborg violates the human/machine distinction; replacing cognition with neural feedback, it challenges the human-animal difference; explaining the behavior of thermostats and people through theories of feedback, hierarchical structure, and control, it erases the animate/inanimate distinction. In addition to arousing anxiety, the cyborg can also spark erotic fascination: witness the female cyborg in Blade Runner. The flip side of the cyborg's violation of boundaries is what Haraway calls its "pleasurably tight coupling" between parts that are not supposed to touch. Mingling erotically-charged violations with potent new fusions, the cyborg becomes the stage on which are performed contestations about body boundaries that have often marked class, ethnic and cultural differences. Especially when it operates in the realm of the Imaginary rather than through actual physical operations (which act as a reality check on fantasies about cyborgism), cybernetics intimates that body boundaries are up for grabs.

As Lakoff and Johnson have shown in their study of embodied metaphors, our images of our bodies, their limitations and possibilities, openings and self-containments, inform how we envision the intellectual territories we stake out and occupy. [4] When the body is revealed as a construct, subject to radical change and redefinition, bodies of knowledge are similarly apt to be seen as constructs, no more inevitable than the organic form that images them. At the same time cybernetics was reconfiguring the body as an informational system, it was also presenting itself as a science of information that would remap intellectual terrains. Branching out into disciplines as different as biology, psychology and electrical engineering, it claimed to be a universal solvent that would dissolve traditional disciplinary boundaries. [5] Wiener, the father of cybernetics, could be supposed to endorse this imperialist ambition. Yet contemplating the penetration of cybernetics into social and humanistic fields, he found himself confronted with some disturbing questions. Where should the cybernetic dissolution of boundaries stop? At what point does the anxiety provoked by dissolution overcome the ecstasy? His writings testify both to the exhilaration and uneasiness that cybernetics generated when its boundary disruptions threatened to get out of hand.

They illustrate the complex dynamics that marked the construction of the cyborg during the foundational period of the late 1040s and 1950s.

As this brief summary suggests, to engage Wiener's work is to be struck by contradiction. Envisioning powerful new ways to equate humans and machines, he also spoke up strongly for liberal humanistic values. A talk given to an audience of physicians in 1954 illustrates the breadth of his concern and ambivalence. [6] He predicted the existence of the automatic factory, argued that electronic computers were thinking machines capable of taking over many human decision-making processes, and cautioned that humans must not let machines become their masters. As I indicated in Chapter 1, the values of liberal humanism--a coherent, rational self, the right of that self to autonomy and freedom, and a sense of agency linked with a belief in enlightened self-interest--deeply inform Wiener's thinking. Often these values stand him in good stead, for example when he rejected lobotomy at a time when Lawrence Kubie, along with many others, was endorsing it. During World War II he frantically immersed himself in military-funded research, but after the war he announced his opposition to nuclear weapons and from then on refused to do military research. [7] The tension between Wiener's humanistic values and the cybernetic viewpoint is everywhere apparent in his writing. On the one hand he used cybernetics to create more effective killing machines (as Peter Galison has noted), applying it self-correcting radar tuning, automated anti-aircraft fire, torpedoes and guided missiles. Yet he also struggled to envision the cybernetic machine in the image of a humanistic self. Ranged alongside his human brother (sisters rarely enter this picture), the cybernetic machine was to be designed so it did not threaten the autonomous, self-regulating subject of liberal humanism. On the contrary, it was to extend that self into the realm of the machine.

But the confluence of cybernetics with liberal humanism was not to run so smoothly. The parallel between self-regulating machinery and liberal humanism has a history that stretches back into the eighteenth century, as Otto Mayr demonstrates in Authority, Liberty, and Automatic Machinery in Early Modern Europe. [8] Mayr argues that ideas about self-regulation were instrumental in effecting a shift from the centralized authoritarian control that characterized European political philosophy during the sixteenth and seventeenth centuries (especially in England, France and Germany) to Enlightenment philosophies of democracy, decentralized control, and liberal self-regulation. Because systems were envisioned as self-regulating, they could be left to work on their own, from the Invisible Hand of Adam Smith's self-regulating market to the political philosophy of enlightened self-interest. These visions of self-regulating economic and political systems produced a complementary notion of the liberal self as an autonomous self-regulating subject. By mid-twentieth century, liberal humanism, self-regulating machinery and possessive individualism had come together in an uneasy alliance that at once helped to create the cyborg and also undermined the foundations of liberal subjectivity. Philip K. Dick tapped into this potential instability when he used his fiction to pose a disturbing question: should a cybernetic machine, sufficiently powerful in its self-regulating processes to become fully conscious and rational, be allowed to own itself? [9] If owning oneself was a constitutive premise for liberal humanism, the cyborg immensely complicate that premise by its figuring of a rational subjectivity that is always already constituted by the forces of capitalistic markets.

The inconsistencies in liberal philosophy that Dick's fiction exposes are also apparent in Wiener's writing: its tendency to use the plural to give voice to a privileged few while presuming to speak for everyone; its masking of deep structural inequalities by enfranchising some while others remain excluded; and its implication in capitalistic imperialism. The closest Wiener comes to a critique of these complicities is a rigid machine he constructs in opposition to the cybernetic machine. This alien and alienating machine is invested with qualities he wants to purge from cybernetics, including rigidity, oppression, militaristic regulation of thought and action, reduction of humans to ant-like elements, manipulation, betrayal, and death. The scope of the critique is limited, for it distances the negative values away from his projects instead of recognizing his complicity with them. When he predicted the automatic factory, for example, he foresaw that it would result in large-scale economic displacements (with all the implications that would have for working class people as autonomous independent agents), but he offered no remedy other than the platitude that men must not let machines take over. [10]

Wiener was not unaware of the ironies through which cybernetics would imperil the very liberal humanistic subject whose origins are enmeshed with self-regulating machinery. Throughout his mature writings, he struggled to reconcile the tradition of liberalism with the new cybernetic paradigm he was in the process of creating. When I think of him, I imagine him laboring mightily to construct the mirror of the cyborg. He stands proudly before this product of his reflection, urging us to look into it so we can see ourselves as control-communication devices, differing in no substantial regard from our mechanical brethren. Then he happens to glance over his shoulder, sees himself as a cyborg, and makes a horrified withdrawal. What assumptions underlie this intense ambivalence? What threads bind them together into something we might call a world view? How are the ambivalences negotiated, and when do they become so intense that the only way to resolve them is to withdraw? What can these complex negotiations tell us about he pleasures and dangers of the posthuman subjectivity that would soon displace the liberal humanist self?

To explore these questions, we will begin with Wiener's early work on probability. In his view, it is because the world is fundamentally probabilistic that control is needed, for the path of future events cannot be accurately predicted. By the same token, control cannot be static or centralized, for then it will not be able to cope with unexpected developments. The necessity for a flexible, self-regulating system of control based on feedback from the system itself starts with the system thumbing its nose at Newtonian predictability. From this, we will follow a web of sticky connections that include a reinscription of homeostasis; a construction of information that grows out of Wiener's deep belief in a probabilistic universe; an interpretation of noise that links it with entropy, degradation, and death; and, above all, an analogical mode of thinking that moves easily across boundaries to identify (or construct) similarities in pattern between very different kinds of structures. As much as anything, it was these analogical moves that helped to construct the cyborg as Wiener envisioned it. All this from a man so uncomfortable with his own body that he could not throw horseshoes in even approximately the right direction and had to abandon a career in biology because he was too clumsy to do the lab work. These physical characteristics are not, I shall argue, entirely irrelevant to the cybernetic viewpoint Wiener was instrumental in forging.

Of Molecules and Men: Cybernetics and Probability

Like Venus, cybernetics was born from the froth of chaos. Wiener's important early work was done on Brownian motion, the random motion molecules make as they collide with each other, bounce off, and collide again, as if they were manic bumper cars. [11] Given this chaos, it is impossible to know the microstates in enough detail to predict from the laws of motion how individual molecules will behave. Therefore probabilistic and statistical methods are required. (The Uncertainty Principle introduced additional complications of a profound nature by setting limits on how precisely positions and momenta can be known.) Probability calculations are facilitated if one assumes that the chaotic motion is homogeneous, that is, the same regardless how the system is sliced to analyze it. Thus the famous ergodic hypothesis: "an ensemble of dynamic systems in some way traces in the course of time a distribution of parameters which is identical with the distribution of parameters of all systems at a given time." [12] Following Birkhoff, Wiener helped to make this hypothesis more limited, precise, and mathematically rigorous than had Willard Gibbs when he first conceived the idea.

Refining Gibbs' methods and ideas, Wiener saw him as a seminal figure not only for his own work, but for all of twentieth century science. "It is . . . Gibbs rather than Einstein or Heisenberg or Planck to whom we must attribute the first great revolution of twentieth century physics," he wrote in The Human Use of Human Beings . [13] Gibbs deserved this honor, Wiener believed, because he realized the deeper implications of probability theory. One explanation for this uncertainty is the limit on knowledge articulated by the Uncertainty Principle, mentioned above. In addition to reflecting our ignorance of microstates, uncertainty also stems from our finitude as human beings. Thirty years before it became an important element in chaos theory, Wiener shrewdly realized that initial conditions can never be known exactly because physical measurements are never completely precise. "What we have to say about a machine or other dynamic system really concerns not what we must expect when the initial positions and momenta are given with perfect accuracy (which never occurs), but what we are to expect when they are given with attainable accuracy" (p. 8).

Related to these epistemological issues is the shift of orientation implicit in Gibbs' approach. Rather than use probabilistic methods to address large numbers of particles (like the bumper cars), Gibbs used probability to consider how different initial velocities and positions might cause a system to evolve in different ways. Thus he considered not many sets within one world but many worlds generated from a single set, or in Wiener's phrase, "all the worlds which are possible answers to a limited set of questions concerning our environment" (Human Use, p. 12). So important did Wiener consider this perspective that he argued "It is with this point of view at its core that the new science of Cybernetics began its development” (p. 12). To see why Wiener considered the innovation profound, we have only to compare it with LaPlace's famous boast that given the initial conditions, a being with enough computing power would be able to predict a system's evolution for eternity. In this view, the universe is completely deterministic and knowable, as precise and predictable as a clock made by God--or what amounted to the same thing for LaPlace, a clock governed by Newton's laws of motion. By contrast, the probabilistic world of Gibbs and Wiener operates like a baggy pair of pants, holding together all right but constantly rearranging itself every time one tries to sit down.

Already steeped in probability theory and inclined to view the world as one evolution realized from a range of possible worlds, Wiener thought about information in the same terms. Working more or less independently of Leon Brillouin and Claude Shannon, he came to similar conclusions. [14] As we saw in Chapter 3, Wiener defined information as a function of probabilities representing a choice of one message from a range of possible messages that might be sent. In a sense, he took Gibbs' idea and substituted word for world. instead of one world coming into being from among a galaxy of possible worlds, one message comes into being from a cacophony of possible messages. When the theory worked, Wiener took it as further confirmation that Gibbs' approach expressed something fundamental about reality; the word and the world are both essentially probabilistic in their natures. This interpretation misses, however, the constructive aspect of information theory. Far from being a passive confirmation, it was an active extension of a probabilistic world view into the new and powerfully synthetic realm of communication theory. We can now understand on a deeper level Wiener's view of cybernetics as a universal theory of knowledge. Such a universal perspective would succeed, he thought, because it reflects the way we know the universe as finite, imperfect creatures. Statistical and quantum mechanics deal with uncertainty on the microscale; communication reflects and embodies it on the macroscale. Envisioning relations on the macroscale as acts of communication was thus tantamount to extending the reach of probability into the social world of agents and actors.

For us in the late age of information, it may seem obvious that communication should be understood as requiring control, and that control should be construed as a form of communication. Underlying this construction, however, is a complex series of events, with its own seriated history of engineering problems, material forms, and bureaucratic structures, that James Beniger has written about in The Control Revolution: Technological and Economic Origins of the Information Society. [15] In broad outline, the forms of control moved from mechanical (a cam directing a mechanical rod to follow a certain path) to thermodynamic (a governor directing the action of a heat engine) to informational (including cybernetic mechanisms of all kinds, from computers to the hypothalamus understood in cybernetic terms). In mechanical exchanges, determinism and predictability loom large. When the center of interest turns to the furnace, with its fiery enactments of Brownian motion, probability necessarily enters the picture. [16] When information comes to the fore, probability moves from being ignorance of microstates to becoming a fundamental attribute of the communication act. As each new form of exchange came to the fore, the older ones did not disappear. An automobile is essentially a heat engine, but it nevertheless continues to use levers and rods of the kind known since the classical era. Similarly, a computer is an information machine, but it also uses molecular processes governed by the laws of thermodynamics. The new forms are distinguished not by the disappearance of the old, but rather by a shift in the control mechanisms they employ, which in turn are determined by the kinds of exchanges the machine is understood to transact.

The move toward cybernetic control theory is itself driven by feedback loops between theory and artifact, research and researcher. Envisioning different kinds of exchanges demanded different kind of control mechanisms, and constructing new control mechanisms facilitated the construction of more exchanges in that mode. [17] The circularity between experimenter, control mechanism, and system interface is part of the story I want to tell. It includes not only the mechanisms of cybernetic systems but also the mindsets of those who constructed themselves and their machines in a cybernetic image. Wiener's assumptions, as we have seen, were rooted in a probabilistic world view. He realized that one of the subtle implications of this view is that messages are constituted, measured, and communicated not as things-in-themselves, but as relational differences between elements in a field. Communication is about relation, not essence.

Across the range of Wiener's writing, the rhetorical trope that figures most importantly is analogy. Understanding communication as relation suggests a deeper reading of this figure. Analogy is not merely an ornament of language but a powerful conceptual mode that constitutes meaning through relation. Seen in this way, analogy is a crucial operator in everything from Wiener's passion for mathematics to his advocacy of "black box" engineering and behaviorist philosophy. Indeed, it is not too much to say that cybernetics as a discipline could not have been created without analogy. When analogy is used to constitute agents in cybernetic discourse, it makes an end run around questions of essence, for objects are constructed through their relations to other objects. Writing in the years immediately preceding and following World War II, Wiener anticipated some aspects of post-structuralist theories. He questioned whether there are any "essential" qualities of humans, animals and machines that exist in themselves, apart from the web of relations that constituted them in discursive and communicative fields. “Whatever view we have of the `realities' underlying our introspections and experiments and mathematical truths is quite secondary; any proposition which cannot be translated into a statement concerning the observable is nugatory,” he wrote in 1936 in “The Role of the Observer.” [18] Wiener also saw sense perception as working through analogy. In his most extreme pronouncement on the matter, he asserted that “Physics itself is merely a coherent way of describing the readings of physical instruments” (a statement deeply regretted by his mathematical biographer, Pesi Masani.) Among the mappings in his view of the world-as-analogy were metaphors that overlaid mathematics onto emotion, sense perception onto communication, and machines onto biological organisms. These mappings throw a different light on his attempts to reconcile cybernetics with a liberal humanistic subject. If meaning is constituted through relation, then juxtaposing men and machines goes beyond bringing two pre-existing objects into harmonious relation. Rather, the analogical relation constitutes both terms through the process of articulating their relationship. To see this meaning-making in process, let us turn now to a consideration of analogy in Wiener's texts and practices.

Crossing Boundaries: Everything Is an Analogy, Including This Statement

In his autobiography I am a Mathematician, Wiener tells of retreating to the family farm for a weekend after a row with a couple of influential Harvard mathematicians. Coming home cold and wet, he falls ill and slips into delirium. "All through the pneumonia," he writes, "my delirium assumed the form of a peculiar depression and worry [about the row and] . . . anxiety about the logical status of my mathematical work. It was impossible for me to distinguish among my pain and difficulty in breathing, the flapping of the window curtain, and certain as yet unresolved parts of the [mathematical] potential problem on which I was working." [19] Retrospectively musing on how his pain merged with external stimuli and mental abstraction, he arrives at a key insight about his relation to mathematics. "I cannot say merely that the pain revealed itself as a mathematical tension, or that the mathematical tension symbolized itself as a pain: for the two were united too closely to make such a separation significant" (p. 85). The realization opens up for him "the possibility that almost any experience may act as a temporary symbol for a mathematical situation which has not yet been organized and cleared up" (p. 86). Identifying an unsolved scientific problem with emotional conflict and physical pain, he becomes "more and more conscious" that for him mathematics serves to "reduce such a discord to semipermanent and recognizable terms" (p. 86). Once he solves the conceptual problem, its link with a personal conflict seems to resolve that as well, allowing him to "release it and pass on to something else" (p. 86). Mapping mathematics onto emotional conflicts is one way, then, that Wiener used analogy. No doubt on more complex grounds than Jacob Bronowski intended, he enthusiastically endorsed Bronowski's suggestion that all of mathematics is a metaphor. Mathematics, he wrote in The Human Use of Human Beings, "which most of us see as the most factual of all sciences, constitutes the most colossal metaphor imaginable, and must be judged, aesthetically as well as intellectually, in terms of the success of this metaphor" (p. 95).

His identification of personal conflicts with conceptual problems was so strong that he perceived it "driving me to mathematics," almost as if against his will (p. 86). The coercive imagery is significant. He was the son of a domineering father who consciously wanted to mold him into a prodigy. Once out from under his father's tutelage, he often found it difficult to motivate himself. Steve Heims, in his biography of Wiener, observes that he apparently used the identification between emotional states and mathematical problems as a spur to goad himself onward. [20] While working on a difficult problem he would fall into a depression, which he would deliberately exacerbate to make himself to work harder. Relying on analogical equivalencies he set up between mathematics and emotion, he anticipated that solving the intellectual problem would allow him to regain psychological homeostasis.

The flip side of drawing analogies is constructing boundaries. Analogy as a figure draws its force from the boundaries it leap-frogs across. Without boundaries, the links it creates would cease to have revolutionary impact. For Wiener, analogy and boundary work went hand in hand. In both his professional and private life, he saw boundaries playing important roles. He included in his first autobiography, Ex-Prodigy: My Childhood and Youth, an account of his mother's anti-Semitism and his feeling of being unwanted and alienated from her when he discovered as a teenager that his father's side of the family was Jewish. [21] Perhaps because of this formative experience, the construction of inside/outside markers characterizes of his response to many life situations. He frequently depicts himself in his autobiographies as an outsider, standing apart from a privileged group whose boundaries do not include him. He made it a point to decline scientific prizes and to resign from prestigious professional groups in which he was offered membership if he did not agree with their goals.

Boundaries also played important roles in his scientific work (as they do in electrical engineering generally). The problem that engaged him when he fell ill and felt the flapping curtain woven into the mathematics was a boundary problem, having to do with what happens to an electrical field around a sharp physical discontinuity. In his later work on cybernetics, boundary formation and analogical linking collaborate to create a discursive field in which animals, humans and machines can be treated as equivalent cybernetic systems. The central text displaying this interplay is the influential cybernetic manifesto that Wiener co-authored with Julian Bigelow and Arturo Rosenblueth, "Behavior, Purpose, and Teleology." [22] Offering an agenda for the nascent field of cybernetics, it also created a discursive style that produced the objects of its analysis.

"Behavior, Purpose and Teleology" begins by contrasting behaviorism with functionalism. Whereas functionalism (in the authors' definition) foregrounds internal structure and is relatively unconcerned with the organism's relation to the environment, behaviorism focuses on relations between the organism and environment and is relatively unconcerned with internal structure. In the laboratory the behaviorist approach leads to "black box" engineering, in which one assumes that the organism is a "black box" whose contents are unknown. Producing equivalent behavior then counts as producing an equivalent system. The obvious justification is that even when little or nothing is known about internal structure, meaningful conclusions can still be drawn about behavior. Bracketing internal structure did more than this, however. It also produced the assertion that because humans and machines sometimes behave similarly, they are essentially alike. Note the slippage in this passage comparing living organisms and machines. "The methods of study for the two groups are at present similar. Whether they should always be the same may depend on whether or not there are one or more qualitatively distinct, unique characteristics present in one group and absent in the other. Such qualitative differences have not appeared so far" (p. 22). "Appeared" is an apt choice of verb, for the behaviorist viewpoint was constructed precisely to elide the very real differences in internal structure that exist between organisms and machines. The analogy is produced by how the focus of attention is constructed. The authors make a similar move when they perform successive cuts in the kinds of behavior they find interesting, focusing, for example, on purposeful rather than random behavior. This series of boundary formations, they contend, "reveals that a uniform behavioristic analysis is applicable to both machines and living organisms, regardless of the complexity of the behavior" (p. 22). What tends to drop from sight is the fact that the equation between organism and machine works because it is seen from a position formulated precisely so it will work.

Another rhetorical move is a reinscription of two important terms, purpose and teleology. Each is carefully defined to fit the cybernetic situation. Purpose implies action directed toward a goal (p. 18); teleology implies a goal achieved through negative feedback. In terms of the offered definitions, teleological behavior means simply "behavior controlled by negative feedback" (p. 24). But keeping a loaded term like teleology in play is not an innocent reinscription. It carries with it the sense of moving toward a goal meaningful to the system that pursues it, which implies that meaning can exist for machines. It also suggests that the behaviorist project has a cosmological dimension appropriate to the sweeping vistas of time and space that teleology is usually taken to imply.

The authors reinforce these implications when they point out that teleology had fallen into scientific disrepute because it posits a "final cause" that exists in time after the effects it is supposed to bring about. Their version of teleology circumvents this problem; it does not rely on Aristotelian causality of any kind, only on purposeful action toward a goal. They suggest that the opposite of teleology is not deterministic causality but non-teleology, that is, random behavior that is not goal-directed. They thus shift onto new ground the centuries-old debate between Newtonian causality and Christian teleology. The important tension now is not between science and God, but between purpose and randomness. Purpose achieved through negative feedback is the way goal-seeking devices deal with a probabilistic universe. By implication, the proper cosmological backdrop for the workings of teleological mechanisms is neither the cosmos infused by divine purpose imagined by Christians or the world of infinite predictability dreamed by LaPlace, but a Gibbesian universe of probabilistic relations and entropic decay. Through these reinscriptions and analogical links, cybernetics becomes philosophy by other means.

A young philosopher, Richard Taylor, took up the gauntlet thrown down in the cybernetic manifesto. In a critique published seven years later in the same journal, The Philosophy of Science, he sought to show that either "purpose" had been stretched so far it could apply to any behavior, or else it had been used to smuggle in inferences about a machine's behavior that properly originated within a human observer. [23] He intended to demonstrate that the rhetoric of "black box" engineering had covertly opened the boxes and put into them qualities produced by the very analysis that treated them as unopened black boxes.

In their rebuttal, Wiener and Rosenblueth make clear that they are appealing to a discourse community of scientists that they deem superior to philosophers. [24] They constitute this community by distinguishing between verbal analysis, which they call "trivial and barren" (p. 318), and their analysis, which is motivated by "scientific" concerns. The implicit contrast between verbal ambiguities that might interest a philosopher and the weighty concerns of "science" is underscored by the contrast they draw between Taylor's "beliefs," and their repeated use of "science" and "scientific" to describe their project (eleven times in a short article). Taylor had used several examples to illustrate that "purpose," as they defined it, could be applied to non-teleological mechanisms (a clock that breaks down at midnight on New Year's Eve, a submarine that follows a boat to which it is attached by a cable). In riposte, Wiener and Rosenblueth contend that these examples are easily distinguished from true servomechanisms using negative feedback. To make the point, however, they are necessarily led into a discussion of the internal structures of the mechanisms--exactly the position they did not want to take in their original article arguing for a behaviorist approach. Their rebuttal is effective, then, only to the extent that it complements a strict behaviorist approach with an analysis that, contra behaviorist principles, uses differences in internal structures to sort behaviors into different categories.

This alternating focus on behavior and internal structure is similar to the rhetorical strategies Geof Bowker analyzes in his article showing how cybernetics constituted itself as a universal science. [25] Bowker points out that cybernetics positioned itself both as a meta-science and as a tool that any other science could use. It offered a transdisciplinary vocabulary that could be adapted for a variety of disciplinary purposes, presenting itself in this guise as content-free, and simultaneously offered a content-rich practice in which cybernetic mechanisms were analyzed, modeled, and occasionally built. Operating at these two different levels, cybernetic discourse was able to penetrate into other disciplines while also maintaining its turf as a disciplinary paradigm. The alternation in Wiener and Rosenblueth's rebuttal between a structure-free and a structure-rich cybernetics produces a similar rhetorical effect. In its structure-free guise, cybernetics links men and machines by eliding internal structure; in its structure-rich form, it presents information flow and negative feedback as important structural elements. It is no accident that Warren McCulloch used a similar rhetorical strategy in his argument with Hans Teuber, discussed in Chapter 3. Just as the alternation between content-free and content-rich cybernetics allows a deeper penetration into disciplinary sites than would otherwise be possible, so the alternation between behavior and structure allowed the discourse simultaneously to assimilate biological organisms and machines into the same category and to distinguish them from plain vanilla mechanical systems.

In his rejoinder, Taylor missed the opportunity to point out that the focus of Wiener and Rosenblueth's analysis alternated between behavior and structure. [26] Instead he chose to pursue a line of questioning similar to his original article, again trying to show that if one relies only on external observations of behavior, "purpose" cannot be reliably distinguished from chance or random events. In contesting for what counts as "purpose," he wanted to deny to the behaviorist approach a distinction crucial to generating their system (the difference between purposeful and random behavior). He sensed that "behavior" had been defined so as to allow intention and desire to be imputed to machines. But he let slip by the larger point that behaviorist assumptions were used selectively to accomplish a political agenda implicit in the way categories were constructed. For Wiener, this agenda included constituting one category that encompassed cybernetic machines and humans, which were put together because they shared the ability to use probabilistic methods to control randomness, and another category for non-cybernetic mechanical systems. These boundary markers implied larger assumptions about the nature of the universe (probabilistic rather than deterministic), effective strategies for dealing with this universe (controlling randomness through negative feedback), and a hierarchy of systems that had moral connotations as well as practical values (flexible systems using negative feedback were better than mechanical devices which did not use feedback). More than the definition of "purpose," it was these larger inscriptions that made "Behavior, Purpose and Teleology" the founding document for cybernetics.

One of the most frequent criticism made of cybernetics during this period was that it was not really a new science, merely an extended analogy (men are like machines). Wiener heard the charge often enough so that he finally felt it was time to take the cybernetic bull by the horns. In a manuscript fragment in the MIT archives entitled "The Nature of Analogy," dated 1950, he offers a strong defense for analogy that moves the argument onto new and more compelling ground. [27] Its brevity notwithstanding, "The Nature of Analogy" is a wide-ranging meditation on what analogy means in science, mathematics, language, and perception. It argues that those who object to Wiener's analogical moves do so because they hold realist assumptions which do not stand up to rigorous scrutiny. Cybernetics as Wiener envisioned it is about relation, not essence. The analogical relations it constructs are therefore not merely rhetorical figures, but systems that generate the only kind of significance available to us as perceiving, finite beings with no access to unmediated reality.

He begins by pointing out that language is always analogical, in the sense that it puts forth propositions the listener must interpret from her experience, which is never identical with the speaker's. The observation anticipates Michael Arbib and Mary Hesse's argument that signification occurs through category constitution, not through the communication of an Aristotelian essence. [28] Like them, Wiener also denies that language communicates an Aristotelian essence. The convergence points to similarities between his definition of information and Saussure's view of la langue, or language as a system. In both cases, communication proceeds through selection from a field of possible alternatives rather than direct articulation of inherent reference. Just as Saussurian linguistics is associated with deconstructive theories that reveal the indeterminacy of reference and expose language's inability to ground itself, so Wiener's cybernetics sees communication as a probabilistic act in a probabilistic universe, where initial conditions are never known exactly and messages signify only through their relation to other messages that might that been sent. For Wiener no less than Saussure, signification is about relation, not about the world as a thing-in-itself. [29]

It is in this context that pattern, associated as we saw in Chapter 2 with information, assumes paramount importance. Wiener's view of sense perception makes the point clear. Perception does not reflect reality directly but rather relies on transformations that preserve a pattern across multiple sensory modalities and neural interfaces. Representation emerges through the analogical relation of these transformations to the original stimulus. In this respect sense perception is like mathematics and logic, for they too "deal preeminently with pattern apart from content" (“Analogy,” p. 2). The behaviorist approach is well suited to this relational epistemology because it concentrates on transmission of patterns rather than communication of essence. Consider the anti-aircraft predictor that Wiener developed in collaboration with Julian Bigelow during World War II. [30] The prognosticator received tracking data as input (for example, radar following a plane) and gave as output predictions of where the plane would go. Statistical analysis was used to find patterns in these data, and the data themselves were understood as patterns analogically related to events in the world. Thus perception, mathematics, and information all concentrate on pattern rather than content. As data moves across various kinds of interfaces, analogical relationships are the links that allow pattern to be preserved from one modality to another. Analogy is thus constituted as a universal exchange system that allows data to move across boundaries. It is the lingua franca of a world (re)constructed through relation rather than grasped in essence.

Border crossings accomplished through analogy include the separation between flesh and world (sense perception), the transition between one discipline and another (for example, moving from the physiology of living organisms to the electrical engineering of a cybernetic machine), and the transformation of embodied experience, noisy with error, into the clean abstractions of mathematical pattern. Even the prostheses Wiener designed can be understood as operating through analogy, for they transformed information from one modality into another. [31] The "hearing glove," for example, was an apparatus that converted sounds (auditory signals) into touch (tactile signals) by stimulating a deaf person's fingers with electromagnetic vibrators that were analogical transformations of sound frequencies. For Wiener, analogy was communication, and communication was analogy. Objecting that cybernetics is "merely an analogy" was for him akin to saying it is "merely about how we know the world."

The problem with this approach lies not so much in the analogical relations Wiener constructed between living and mechanical systems, as in his tendency to erase from view the very real differences in embodied materiality that the analogies did not express. Confronted with two situations, he was much more inclined to move easily and quickly to an abstract level where similarities in patterns became evident than to remain attentive to the particularities that made each situation unique. No doubt his own lack of involvement in the nitty-gritty work of the lab was a contributing factor in this elision of embodied materiality. In his autobiography, he wrote about the impatience he felt with the exacting procedures of the biological laboratory. "This impatience was largely the result of my mental quickness and physical slowness. I could see the end to be accomplished long before I could labor through the manipulative stages that were to bring me there." [32] The problem was serious enough to force him to give up his hope of earning a Ph.D. in biology. In his later professional collaborations with Rosenblueth and others, he left the lab work to them. Colleagues recall how he would wander into Rosenblueth's laboratory when an experiment was underway, make a few notes and ask a few questions, and retreat to his office to work out the mathematical analogies expressing the physical situation. When Wiener and his collaborators wrote such phrases as "We cut the attachment of the muscle," the plural was purely honorary, as Masani points out in his excellent biography. [33] Other colleagues, particularly Walter Rosenblith and Jerome Wiesner, suggested in a posthumous tribute to Wiener that his ineptitude in the lab made him less attentive than they would have wished to the particularities of actual neurophysiological structures. "In areas in which Wiener's intuition was less educated than in engineering, he was often impatient with experimental details; for example, he seemed sometimes unwilling to learn that the brain did not behave the way he expected it to.” [34] For Wiener, the emphasis on analogy went hand in hand with a certain estrangement from the flesh. In this respect, the contrast between him and McCulloch is clear. As a dedicated experimentalist, McCulloch was sensitive in a way that Wiener was not to the tension between the plenitude of embodiment and the sparseness of abstraction.

As we have seen, Wiener wanted to inscribe cybernetics into a larger drama that would reinforce the liberal humanistic subject. Given his inclination toward a Gibbesian universe, that drama focused on probability. In addition to operating on the microscale of subatomic particles and the macroscale of cybernetic circuits, probability also operates on the cosmological level of universal dissipation and decay. Linking probability with information allowed Wiener to script the cybernetic subject into a cosmological drama of chaos and order. It is here, on this cosmological level, that he stages the moral distinctions between good cybernetic systems which reinforce the autonomous liberal subject, and evil machines which undermine or destroy the autonomy of the subject. An important player in this titanic struggle between good and evil machines is entropy, a protean concept with a richly complex history. Let us turn now to consider this key actor.

Entropy as Cultural Relay: From Heat Engines to Information

We can begin our investigation into entropy with the series of transformations that Mark Seltzer traces in Bodies and Machines. [35] Seltzer, concentrating on the social formations of late nineteenth-century naturalism, finds at the heart of naturalism a double and seemingly contradictory thrust: on the one hand "the insistence on the materiality or physicality of persons, representations, and actions"; on the other hand, "the insistent abstraction of persons, bodies, and motions to models, numbers, maps, charts, and diagrammatic representations" (p. 14). Calling the ideology that resulted from this double thrust a "dematerialized materialism" (p. 14), Seltzer instances such phenomena as the emergence of statistical representations for human behavior and a renewed interest in the ergonomics of the human body. One focuses on behavior abstracted into statistical ensembles of data, the other on the material processes of energy consumption and dissipation. They illustrate the construction of bodies both as material objects and as probability distributions.

The duality Seltzer locates in late nineteenth-century culture continued into the twentieth century with renewed force when statistical thermodynamics merged with information theory. One of the principal sites for this merger was cybernetics. The emphasis on pattern constructed bodies as immaterial flows of information; the alternating emphasis on structure recognized that these "black boxes" were heavy with materiality. Complex couplings between the two registers worked to set up a series of exchanges between biological organisms and machines. To see how these couplings evolved, let us start with the exchanges that thermodynamics set up and follow them forward into cybernetics.

The first law of thermodynamics, stating that energy is neither created or destroyed, points to a world in which no energy is lost. The second law, stating that entropy always tends to increase in a closed system, forecasts a universe that is constantly winding down. This tension between the first and second laws, stability and degradation, runs like a leitmotif through turn-of-the-century cultural formations. According to Seltzer the tension itself acts like a thermodynamic exchanger, allowing incompatible terms such as production and reproduction, machines and bodies, to be articulated together. The body is like a heat engine because it cycles energy into different forms and degrades it in the process; the body is not like a heat engine because it can use energy to repair itself and to reproduce. In one sense the comparison constructs the difference between body and machine; in another sense it acts as an exchanger that allows bodies and heat engines to be linked together. Through such comparisons, Seltzer argues, "what is gradually elaborated is a more or less efficient, more or less effective system of transformations and relays between 'opposed' and contradictory registers" (p. 41). These ambiguous linkages were reinforced because thermodynamics itself was perceived as operating in the two different registers of conservation and dissipation. Thus he concludes that thermodynamics, wrapping both conservative stability and dissipative decay within the mantle of scientific authority, "provided a working model of a new mechanics and biomechanics of power" (p. 41).

Already functioning as an exchange system within the culture, thermodynamics evolved into "dematerialized materialism" when Boltzmann gave entropy a much more general formulation by defining it as a probability function. In this "dematerialized" construction, entropy was interpreted as a measure of randomness. The second law was then re-formulated to state that closed systems tend to move from order to randomness. Encompassing entropy's earlier definition, Boltzmann's formulation also added something new, for it allowed entropy to be linked with systems that had nothing to do with heat engines.

This dematerialization was carried further when entropy was connected with information. As early as 1929, the connection was made through Leo Szilard's interpretation of Maxwell's Demon, [36] a mythical being in a thought experiment proposed by James Clerk Maxwell in 1871. The demon gained energy by sorting molecules. Szilard and Leon Brillouin, among others, pointed out that to sort molecules, the demon has to have information about them. [37] The container in which the demon sits is imagined as a “black body” (a technical term meaning that the radiation is uniformly dispersed), so there is no way for the demon to "see" the molecules. Brillouin calculated that the energy the demon would have to expend to get information about the molecules is greater than what he could gain by the sorting process. The immediate result was to rescue the second law, which in any case was too well-established to be seriously in doubt. The more important implication was to suggest that entropy and information are inversely related to each other. The more information, the less entropy; the more entropy, the less information. Brillouin therefore proposed that information be considered as negative entropy, or negentropy. Maxwell's Demon was one of the relay points through which a relationship was established between entropy and information.

Like Brillouin and many others of his generation, Wiener accepted the idea that entropy was the opposite of information. The inverse relation made sense to him because he thought of information as allied with structure, and entropy with randomness, dissipation, and death. "As entropy increases," he wrote, "the universe, and all closed systems in the universe, tend naturally to deteriorate and lose their distinctiveness, to move from the least to the most probable state, from a state of organization and differentiation in which distinctions and forms exist, to a state of chaos and sameness. In Gibbs' universe order is least probable, chaos most probable" (p. 12). In this view, life is an island of negentropy amidst a sea of disorder. "There are local enclaves whose direction seems opposed to that of the universe at large and in which there is a limited and temporary tendency for organization to increase. Life finds its home in some of these enclaves" (p. 12). In a related metaphor, he envisioned a living organism as an informational system swimming upstream against the entropic tide.

This view of entropy makes sense when viewed in the context of nineteenth-century thermodynamics. But it is not a necessary implication of information as it is technically defined. Claude Shannon took the opposite view and identified information and entropy rather than opposed them. [38] Since the choice of sign was conventional, this formulation was also a possibility. Heuristically, Shannon's choice was explained by saying that the more unexpected (or random) a message is, the more information it conveys. [39] This change in sign did not affect the dematerialization entropy had undergone, but it did reverse its value in more than a mathematical sense. In retrospect, identifying entropy with information can be seen as a crucial crossing point, for it allowed entropy to be re-conceptualized as the thermodynamic motor driving systems to self-organization rather than the heat engine driving the world to universal heat death. Space will not permit me to tell the story of this reversal here, and in any event, it has been chronicled elsewhere. [40] Suffice it to say that as a result, chaos went from being associated with dissipation in the Victorian sense of dissolute living and reckless waste to a newly positive sense that associated dissipation with increasing complexity and new life.

Wiener came close to making this crossing. In one of his astonishing analogical leaps, he saw a connection between the "light" the Demon needs to sort the molecules and the lights plants use in photosynthesis. He argued that in photosynthesis, plants act as if their leaves were studded with Maxwell's Demons, all sorting molecules to allow the plant to run uphill toward increasing complexity rather than downhill towards death. [41] But he did not go beyond this isolated insight to the larger realization that large entropy production could drive systems to increasing complexity. Finally he remains on the negative side of this divide, seeing life and homeostasis as contrarian islands that, although they may hold out for a while, must eventually be swamped by the entropic tide.

So firmly rooted is Wiener in this perspective that he comes close on several occasions to saying that entropic decay is evil. Entropy becomes morally negative for Wiener when he sees it operating against the differential probability distributions on which the transfer of information depends. Recall that Gregory Bateson defined information as a difference that makes a difference; if there is no difference, there is no information. Since entropy tends always to increase, it will eventually result in a universe where all distributions are in their most probable state and universal homogeneity prevails. Imagine Dr. Zhivago sitting at his desk in a cold, cold room, trying to telegraph a message to his beloved Laura, while in the background Laura's theme plays and entropy keeps relentlessly increasing. Icicles hanging from his fingers and the telegraph key, he tries to tap out "I love you," but he is having trouble. Not only is he freezing from heat death; he is also stymied by information death. No matter what he taps, the messages always come out the same: eeeeeee (or whatever letter is most common in the Russian alphabet). This whimsical scenario illustrates why Wiener associated entropy with oppression, rigidity, and death. Communication can be seen, he suggested, as a game that two men (or machines) play against noise. [42] To be rigid is inevitably to lose the game, for it consigns the players to the mechanical repetition of messages that can only erode over time as noise intervenes. Only if creative play is allowed, if the mechanism can adapt freely to changing messages, can homeostasis be maintained, even temporarily, in the face of constant entropic pressure toward degradation.

In the "dematerialized materialism" of the battlefield where life struggles against entropy and noise, the body ceases to be regarded primarily as a material object and instead is seen as an informational pattern. The struggle, then, is between strategists who try to preserve this pattern intact, and noisy opponents (or rather, noise as an opponent) who try to disrupt it. Wiener was one of the important voices during the 1940s and 1950s who cast the cosmological drama between cybernetic mechanisms and noise in these terms. In The Human Use of Human Beings, he suggests that human beings are not so much bone and blood, nerve and synapse, as they are patterns of organization. He points out that over the course of a lifetime the cells comprising a human being change many times over. Identity cannot therefore consist in physical continuity. "Our tissues change as we live: the food we eat and the air we breathe become flesh of our flesh and bone of our bone, and the momentary elements of our flesh and bone pass out of our body every day with our excreta. We are but whirlpools in a river of ever-flowing water. We are not stuff that abides, but patterns that perpetuate themselves" (p. 96). Consequently, to understand humans it is necessary to understand how the patterns of information they embody are created, organized, stored, and retrieved. Once these mechanisms are understood, they can be used to create cybernetic machines. If memory in humans is the transfer of informational patterns from the environment to the brain, machines can be built that will effect the same kind of transfer. Even emotions may be achievable for machines, if feelings are considered not as "merely a useless epiphenomenon of nervous actions" (p. 72) but as control mechanisms governing learning. [43] Considered as informational patterns, cybernetic machines and men can make common cause against the disruptive forces of noise and entropy.

The picture that emerges from these conjectures shows the cybernetic organism--human or mechanical--responding flexibly to changing situations, learning from the past, freely adapting its behavior to meet new circumstances, and succeeding in preserving homeostatic stability in the midst of even radically altered environments. Nimbleness is an essential weapon in this struggle, for to repeat mindlessly and mechanically is inevitably to let noise win. Noise has the best chance against rote repetition, where it goes to work at once to introduce randomness. But a system that already behaves unpredictably cannot be so easily subverted. If a Gibbesian universe implies eventual information death, it also implies a universe where the best shot for success lies in flexible and probabilistic behavior. The Greek root for cybernetics, 'steersman,' aptly describes the cybernetic man/machine: light on his feet, sensitive to change, himself a flow and knowing how to go with the flow.

Reinforcing the boundary work that assimilates the liberal humanistic subject and cybernetic machine into the same privileged space are the distinctions Wiener makes between good and bad machines. When machines are evil in The Human Use of Human Beings, it is usually because they have become rigid and inflexible. Whereas the cybernetic machine is ranged alongside man as his brother and peer, metaphors that cluster around the rigid machine depict it through tropes of domination and engulfment. The ultimate horror is for the rigid machine to absorb the human being into itself, co-opting the flexibility that is the human birthright. "When human atoms are knit into an organization in which they are used, not in their full right as responsible human beings, but as cogs and levers and rods, it matters little that their raw material is flesh and blood. What is used as an element in a machine, is in fact an element in the machine" (p. 185). Here the analogical mapping between man and machine turns sinister, trapping man within inflexible walls that rob him of his autonomy. The passage shows how important it is to Wiener to construct the boundaries of the cybernetic machine so that it reinforces rather than threatens the autonomous self. When the boundaries turn rigid or engulf humans so they lose their agency, the machine ceases to be cybernetic and becomes simply and oppressively mechanical.

The cosmological stage upon which the struggle between oppressive machines and cybernetic systems unfolds is--no surprise--the Gibbesian universe where probability reigns supreme. "The great weakness of the machine--the weakness that saves us so far from being dominated by it--is that it cannot yet take into account the vast range of probability that characterizes the human situation" (p.181). Here the probability differentials that make communication possible are assimilated to humans and good machines, leaving bad machines to flounder around in probabilities too diverse for them to assess. The rules of the contest are laid down by the second law of thermodynamics, which allows a margin in which cybernetic men/machines can operate because it is still cranking up its death engine. "The dominance of the machine presupposes a society in the last stages of increasing entropy, where probability is negligible and where the statistical differences among individuals are nil. Fortunately we have not yet reached such a state" (p. 181). When in the end the universe ceases to manifest diverse probabilities and becomes a uniform soup, control, communication, cybernetics--not to mention life--will expire. In the meantime, men and cybernetic machines stand shoulder to shoulder in building dikes that temporarily stave off the entropic tide.

The boundary work that links cybernetic machines and humans perhaps reaches its most complex articulation in the distinction Wiener makes between Augustinian and Manichean opponents. At issue in this distinction is the difference between an opponent who plays "honorably," that is, by abiding rules that do not change, and one who tries to win by manipulating and psyching out his opponent. The exemplar of an Augustinian opponent for Wiener is nature. Nature--including noise--may sometimes frustrate the scientist's attempt to control it, but it does not consciously try to manipulate its opponent. The exemplar of the Manichean opponent is the chess-player, including chess-playing machines. Unlike nature, the chess-player acts deviously and, if possible, manipulatively. When the chess player is contrasted with the scientist, it is almost always to the chess player's detriment. In pointing out that nature does not try to outwit the scientist, Wiener observes that having an Augustinian opponent means the scientist has time to reflect and correct his strategy, because no one is trying to take advantage of his mistakes. Whereas the scientist is thus governed by his best moments, the chess player is governed by his worst (p. 36).

Peter Galison, in "The Ontology of the Enemy: Norbert Wiener and the Cybernetic Vision," argues that cybernetics (along with game theory and operations research) should be called a "Manichean science.” [44] In a fine-grained analysis of Wiener's collaboration with Julian Bigelow to develop an anti-aircraft (AA) weapon during World War II, Galison brilliantly shows that Wiener did not construct the enemy as a sub-human rapacious predator, as was so often the case in war propaganda. Rather, he constructed the enemy--for example, a fighter pilot trying to evade AA fire--as a probabilistic system that could effectively be countered using cybernetic modeling. Unlike other fire systems that had fixed rules derived from probabilistic modeling, Wiener imagined a firing machine that could evolve new rules based on prior observation--imagined, that is, a machine that could learn. Thus the firing system would evolve to become as Manichean as the enemy it faced. Galison argues that this strategy enabled a series of substitutions and identifications to take place that mapped the enemy pilot onto the servo-controller and ultimately onto the allied war personnel behind the servo-controller. In a "Summary Report for Demonstration," Wiener and Bigelow wrote, "We realized that the 'randomness' or irregularity of an airplane's path is introduced by the pilot; that in attempting to force his dynamic craft to execute a useful manoeuvere . . . the pilot behaves like a servo-mechanism “ (quoted in Galison, p. 236). Thus cybernetics, itself constituted through analogies, creates further analogies through theories and artifacts that splice man to machine, German to American. Through this relay system, the enemy becomes like us and we like him: enemy mine. If these analogical mappings relay kept the enemy pilot being demonized, they also made the cybernetic machine (and by extension, cybernetics itself) party to a bloody struggle in which Manichean tactics were used by both sides to kill as many humans as possible.

It was partly in reaction against this cooptation of cybernetics by the military that Wiener, half a decade after the war, wrote the significantly entitled The Human Use of Human Beings. [45] Although Wiener had done everything in his power during the war to further cybernetics as a “Manichean science,” his writings, especially after the war, show a deep aversion to the manipulation that a Manichean opponent implies. From his autobiographies, it is clear that he was hyper-sensitive to being manipulated, perhaps with good reason. When he was first beginning to establish himself as a mathematician, his father tried to get him to use his contacts to advance the father's philological ideas--an instance of manipulation that made Wiener increasingly wary of how others might try to use his talents and influence to further their own ends. It is no accident that he associates the manipulative chess-playing machine with the military projects that he turned resolutely away from after atomic bombs vaporized hundreds of thousands of Japanese civilians. Remarking on Claude Shannon's suggestion that chess-playing machines have military potential, he writes, "When Mr. Shannon speaks of the development of military tactics, he is not talking moonshine, but is discussing a most imminent and dangerous contingency" (Human Use, p. 178). The problem, of course, was that cybernetics adapted all too readily to Manichean tactics, making it possible to play these deadly games even more effectively.

Wiener's war work, combined with his anti-military stance after the war, illustrate with startling clarity how cybernetics functioned both as a source of intense pride and intense anxiety for him. This tension, often expressed as an anxious desire to limit the scope of cybernetics, takes a different but related form when he considers the question of body boundaries, always a highly charged issue. When the physical boundaries of the human form are secure, he celebrates the flow of information through the organism. All this changes, however, when the boundaries cease to define an autonomous self, either through manipulation or engulfment. In the next section, we will see how this anxiety erupts into the 1948 Cybernetics at critical points, causing him to withdraw from the more subversive implications of the discipline he fathered. It is no accident that erotic metaphors are used to carry the argument's thrust. Like cybernetics, eroticism is intensely concerned with the problematics of body boundaries. It is not for nothing that sexual orgasm is called "the little death" or that writers from Marquis de Sade to J. G. Ballard have obsessively associated eroticism with penetrating and opening the body. At stake in the erotically charged discourse where Wiener considers the pleasures and dangers of coupling between parts that are not supposed to touch is how extensively the body of the subject may be penetrated or even dissolved by cybernetics as a body of knowledge. It is here, as much as anywhere, that Wiener's concern to preserve the liberal subject comes into uneasy tension with his equally strong desire to advance the cause of cybernetics. That resolution can be achieved only by withdrawal points toward a future in which the cybernetic subject could finally not be contained within the assumptions of liberal humanism.

The Argument for Celibacy: Preserving the Boundaries of the Subject

In Cybernetics , the technical text from which The Human Use of Human Beings was adapted, Wiener looks into the mirror of the cyborg but then withdraws. [46] The scenarios he constructs to enact and justify this withdrawal point suggestively to the role that erotic anxiety plays in cybernetic narratives. In my analysis I will focus on the chapter entitled "Information, Language, and Society." Here Wiener entertains the possibility that cybernetics has provided a way of thinking so fertile that it will allow the social and natural sciences to be synthesized into one great field of inquiry. Yet he finally demurs from this palpable object of desire. Given that he is as imperialistic as most scientists who think they have invented a new paradigm, why does he prefer to maintain the intellectual celibacy of his discovery? I will argue that central to his decision is a fantasy scene that expresses and controls anxiety by reconstituting boundaries. This fantasy gives rise to a series of encoded metaphors that appear whenever anxiety becomes acute. The metaphors also have literal meanings that reveal how intermingled the physical remains with the conceptual, the erotic with the cybernetic. As gestures of separation disconcertingly transform into couplings, the cybernetics of the subject and the subject of cybernetics interpenetrate.

Wiener works up to the fantasy by pointing out that there are many organizations whose parts are themselves small organizations. Hobbes's Leviathan is a Man-State made up of men; a Portuguese man-of-war is made up of polyps that mirror it in miniature; a man is an organism made of cells that in some respects also function like organisms. This line of thought leads him to ask how these "bodies politic" function. "Obviously, the secret is in the intercommunication of its members" (p. 156). The flow of information is thus introduced as a principle explaining how organization occurs across multiple hierarchical levels. To illustrate, he instances the "sexually attractive substances" that various species secrete to insure that the sexes will be brought together (p. 156). For example, the pheromones that guide insect reproduction are general and omnidirectional, acting in this respect like hormones secreted within the body. The analogy suggests that external hormones organize internal hormones, so that a human organism becomes, in effect, a sort of permeable membrane through which hormonal information flows. At this point we encounter his first demurral. "I do not care to pronounce an opinion on this matter," he announces rather pretentiously after introducing it, preferring to "leave it as an interesting idea" (p. 157).

I think that the idea is left because it is disturbing as well as speculative. It implies that personal identity and autonomous will are merely illusions that mask the cybernetic reality. If our body surfaces are membranes through which information flows, who are we? Are we the cells that respond to the stimuli? The larger collective whose actions are the resultant of individual members? Or the host organism who, as Richard Dawkins was later to claim using cybernetic arguments, engages in sex because it is controlled by selfish genes within? [47] The choice of examples foregrounds sexuality, but this is a kind of sex without sexuality. Implying the deconstruction of the autonomous self as a locus of erotic pleasure, it circumvents the assenting, demurring, intensifying, delaying, and consummating that constitute sexual play. Confronted with this sexless sex, Wiener's first impulse is to withdraw: coitus interruptus.

His second impulse is to reconstruct himself as a liberal subject through a disguised erotic fantasy that allows him to control the flow of information, rather than be controlled by it. Similar fantasies appear everywhere in American literature, from Natty Bumppo and Chingachgook to Ishmael and Queequeg. They are ubiquitous because they are about the American values of masculine autonomy and control, about deferred intimacy between men in a society that is homophobic, racist, and misogynist. What is this fantasy? What else but for the American male to imagine himself alone in the woods with an "intelligent savage," giving himself over to the pursuits men follow when they are alone together (p. 157).

The fantasy's ostensible purpose is to show that Wiener and his savage companion could achieve intimacy, even if they did not touch and shared no language. Wiener imagines himself "alert to those moments when [the savage] shows the signs of emotion or interest," noticing at these moments what he watches (p. 157). After a time, the savage would learn to reciprocate by "pick[ing] out the moments of my special, active attention," thus creating between them "a language as varied in possibilities as the range of impressions that the two of us are able to encompass" (p. 157). Alone together in the woods, the two men construct a world of objects through the interplay of their gazes. In the process they also reconstitute themselves as autonomous subjects who achieve intimacy through their voyeuristic participation in each other's emotion and "special, active attention." There remains of course a necessary difference between them. Wiener can move from this fantasy to the rest of his argument, whereas the "intelligent savage" reappears in his discourse only when he finds it convenient to invoke him. The passage reveals in miniature how the use of the plural by the liberal humanist subject can appropriate the voices of subaltern others, who if they could speak for themselves, might say something very different.

Having reassured himself of intimacy, autonomy and control, Wiener returns to the problem of the "body politic," concentrating on its alarming lack of homeostasis. In contrast to the regulated, orderly exchanges between him and his savage friend, the body politic is dominated by exchanges between knaves and fools, with "betrayal, turncoatism, and deception" the order of the day (p. 59). The economy of this society is clear-cut: the fools desire, the knaves manipulate their desires. The economy is reinforced by statisticians, sociologists, and economists who prostitute themselves by figuring out for the knaves exactly how the calculus of desire can be maximized. The only respite from this relentless manipulation are small, autonomous populations. There homeostasis can still work, whether in "highly literate communities. . . or villages of primitive savages" (p. 160). The reappearance of the savage here is significant, for it occurs just when anxiety about having desire manipulated reaches its height. No doubt it has a soothing effect on Wiener's imagination, for it reminds him that after all he need not be manipulated.

We come now to the crux of the argument. The danger of cybernetics, from his point of view, is that it can potentially annihilate the liberal subject as the locus of control. On the microscale, the individual is merely the container for still smaller units within, who dictate its actions and desires; on the macroscale, these desires make one into a fool to be manipulated by knaves. Under a cybernetic paradigm, these two scales of organization would be joined to each other. What chance then for intimate communication alone with an intelligent savage in the woods? No, despite the "hopes which some of my friends have built for the social efficacy of whatever new ways of thinking this book may contain," he finds himself unable to attribute "too much value to this type of wishful thinking" (p. 162). Ironically, expanded too far across the bodies of disciplines, the science of control might rob its progenitor of the very control that was no doubt for him one of its most attractive features.

Having reached this conclusion, he reenacts the anxiety that gave rise to it through a series of interactive metaphors that connect his fantasy with his anxiety. He claims that it is a "misunderstanding of the nature of all scientific achievement" to suppose that "the physical and social sciences can be joined" (p. 162). They must be kept apart, for they permit different degrees of coupling between the scientist and the object of his interest. The precise sciences "achieve a sufficiently loose coupling with the phenomena we are studying [to allow us] to give a massive total account of this coupling" (p. 163). Erotic interest is not altogether lacking, for "the coupling may not be loose enough for us to be able to ignore it altogether" (p. 163). Nevertheless, the restrained science he practices is of a different order than the social sciences, where the coupling is much tighter and more intense. The contrast shows how central the concept of the autonomous self is to cybernetics as Wiener envisioned it.

The savage makes one last appearance in Wiener's anxious consideration of how tightly the scientist can be coupled with his object without losing his objectivity. To illustrate the dangers of tight coupling, he observes that primitive societies are very often changed by the anthropologists who observe them. He makes the point specifically in terms of language, noting that "Many a missionary has fixed his own misunderstandings of a primitive language as law eternal in the process of reducing it to writing" (p. 63). In implicit contrast to this violation is the pristine intimacy he achieved with his savage, where no misunderstandings disrupted the perfect sympathy of their gazes.

Concluding that "We are too much in tune with the objects of our investigations to be good probes," he counsels that cybernetics had best be left to the physical sciences, for to carry it into the human sciences would only build "exaggerated expectations" (p. 164). Behind this conclusion is the prospect of an interpenetration so complete that it would link the little units within to the larger social units without, thereby reducing the individual to a connective membrane with no control over his desires and no ability to derive pleasure from them. Not only sex, but the sex organs themselves, disappear in this construction. Thus he decides that, however tempting the prospect of penetrating the boundaries of other disciplines might be, cybernetics is better off remaining celibate.

The conjunction of erotic anxiety and intellectual speculation in Wiener's text implies that cybernetics cannot be adequately understood simply as a theoretic and technological extension of information theory. The analogies so important to his thought are constituted not only through similarities between abstract forms (such as probability ratios and statistical analysis), but also through the complex life world of embedded physicality that natural language expresses and evokes through its metaphoric resonances. Natural language is not extraneous to understanding the full complexities of Wiener's thinking, as his mathematical biographer Pesi Masani implies when he contrasts the disembodied abstractions of mathematics with the "long-winded verbosity [of natural language], the hallmark of bureaucratic chicanery and fake labor" (p. 21). On the contrary, the embodied metaphors of language are crucial to understanding the ways in which Wiener's construction of the cybernetic body and the body of cybernetics both privilege and imperil the autonomous humanistic subject.

Viewed in historical perspective, Wiener was not successful in containing cybernetics within the circle of liberal humanistic assumptions. Only for a relatively brief period in the late 1940s and 1950s could the dynamic tension between cybernetics and the liberal subject be maintained, uneasy and anxious as that accommodation often was for Wiener. By the 1960s, the link between liberal humanism and self-regulation that had been forged in the eighteenth century was already stretched thin; by the 1980s, it was largely broken. It is to Wiener's credit that he tried to craft a version of cybernetics that would enhance rather than subvert human freedom. But no person, even the father of a discipline, can control single-handedly what it signifies when it propagates through the culture by all manner of promiscuous couplings. Even as cybernetics lost the momentum of its drive to be a universal science, its enabling premises were mutating and reproducing at other sites. The voices that speak the cyborg do not speak as one, and the stories they tell are very different narratives than those Wiener struggled to authorize.




[2]Gregory Bateson, Steps to an Ecology of Mind (New York: Ballantine Books, 1972).

[3]Donna Haraway, "Manifesto for Cyborgs: Science, Technology, and Socialist Feminism in the 1980s," Socialist Review 80 (1985): 65-108.

[4]George Lakoff and Mark Johnson, Metaphors We Live By (Chicago: University of Chicago Press, 1980); and Mark Johnson, The Body in the Mind: The Bodily Basis of Meaning, Imagination, and Reason (Chicago: University of Chicago Press, 1987).

[5]For an analysis of the strategies used to proclaim cybernetics a universal science, see Geof Bowker, "How To Be Universal: Some Cybernetic Strategies, 1943-1970," Social Studies of Science 23 (1993): 107-27.

[6]Norbert Wiener, "Men, Machines, and the World About," box 13, folder 750, Norbert Wiener Papers, Collection MC-22, Institute Archives and Special Collections, Massachusetts Institute of Technology Archives, Cambridge MA. Hereafter called NWP. Also published in Norbert Wiener: Collected Works with Commentaries, edited by Pesi Masani, vol. IV (Cambridge: MIT Press, 1985), pp. 793-799.

[7]For a study tracing Wiener's post-war views, see Steve J. Heims, John von Neumann and Norbert Wiener: From Mathematics to the Technologies of Life and Death (Cambridge: MIT Press, 1980).

[8]Otto Mayr, Authority, Liberty, and Automatic Machinery in Early Modern Europe (Baltimore: Johns Hopkins University Press, 1986).

[9]The question is posed most powerfully in Philip K. Dick, Blade Runner (originally published under the title Do Androids Dream of Electric Sheep?) (New York: Ballantine Books, 1982; orig. pub. date 1968).

[10]It remained for a novelist, Kurt Vonnegut, to envision the fully implications of Wiener's cybernetic program if it were fully carried out in Player Piano (New York: Delacorte Press/Seymour Laurence, 1952).

[11]See for example Norbert Wiener, "The Averages of an Analytical Function and the Brownian Movement," Coll. Works I, pp. 450-455.

[12]Norbert Wiener, "The Historical Background of Harmonic Analysis," American Mathematical Society Semicentennial Publications Vol. II, "Semicentennial Addresses," American Mathematical Society, (Providence, R.I.: 1938), pp. 513-522.

[13]Norbert Wiener, The Human Use of Human Beings (New York: Anchor Books, 1954. org. pub., Boston: Houghton Mifflin, 1950), p. 10.

[14]Wiener of course knew Shannon; both were participants in the Macy conferences. Although they conceived of information in similar ways, Wiener was more inclined to see information and entropy as opposites.

[15]James R. Beniger, The Control Revolution: Technological and Economic Origins of the Information Society (Cambridge: Harvard University Press, 1986).

[16]Michel Serres brilliantly analyzes the progression from the mechanical to the thermodynamical in "Turner Translates Carnot," Hermes: Literature, Science, Philosophy, edited by Josué V. Harari and David F. Bell (Baltimore: Johns Hopkins University Press, 1982).

[17]Beniger convincingly shows how technologies of speed and communication precipitated a "crisis of control" that, once solved, initiated a new cycle of crisis (1986).

[18]Norbert Wiener, “The Role of the Observer,” Philosophy of Science 3 (1936): 307-319, especially p. 311.

[19]Norbert Wiener, I Am a Mathematician: The Later Life of a Prodigy (Garden City, NY: Doubleday, 1956), p. 85.

[20]Heims, John von Neumann and Norbert Wiener, pp. 155-57.

[21]Norbert Wiener, Ex-Prodigy: My Childhood and Youth (New York: Simon and Schuster, 1953).

[22]Arturo Rosenblueth, Norbert Wiener and Julian Bigelow, "Behavior, Purpose and Teleology," Philosophy of Science 10 (1943): 18-24.

[23]Richard Taylor, "Comments on a Mechanistic Conception of Purposefulness," Philosophy of Science 17 (1950): 310-317.

[24]Arturo Rosenblueth and Norbert Wiener, "Purposeful and Non-Purposeful Behavior," Philosophy of Science 17 (1950): 318-326.

[25]Geof Bowker, “How to Be Universal,” pp. 107-27.

[26]Richard Taylor, "Purposeful and Non-Purposeful Behavior: A Rejoinder," Philosophy of Science 17 (1950): 327-332.

[27]Norbert Wiener, "The Nature of Analogy," (1950), box 12, folder 655, NWP.

[28]Michael A. Arbib and Mary B. Hesse, The Construction of Reality (Cambridge: Cambridge University Press, 1986).

[29]Michael J. Apter draws the comparison between Sassurian linguistics and cybernetics in "Cybernetics: A Case Study of a Scientific Subject: Complex," The Sociology of Science: Sociological Review Monograph No. 18, edited by Paul Halmos (Keele, Staffordshire: Keele University, 1972), pp. 93-115, especially p. 104.

[30]I rely here on Peter Galison's detailed account of Wiener's work with AA, “The Ontology of the Enemy: Norbert Wiener and the Cybernetic Vision,” Critical Inquiry 21 (1994): 228-266. (Galison, 1994).

[31]Norbert Wiener, "Sound Communication with the Deaf," Coll. Works IV, pp. 409-11.

[32]Cited in Walter A. Rosenblith and Jerome B. Wiesner, "From Philosophy to Mathematics to Biology: The Life Sciences and Cybernetics," "Norbert Wiener," 3-8, especially 3.

[33]Pesi Masani, Norbert Wiener, 1894-1964 , Vita Mathematica Series, vol. 5 (Basel: Birkhaeuser, 1989), pp. 205-06.

[34]Rosenblith and Wiesner, p. 7.

[35]Mark Seltzer, Bodies and Machines (New York and London: Routledge, 1992).

[36]Leo Szilard, "On the Reduction of Entropy as a Thermodynamic System Caused by Intelligent Beings," Zeitschrift für Physik 53 (1929): 840-856.

[37]Leon Brillouin, "Maxwell's Demon Cannot Operate: Information and Entropy, I," Journal of Applied Physics, 212 (March 1951): 334-357. Much of this material is also available in Maxwell's Demon: Entropy, Information, Computing, edited by Harvey S. Leff and Andrew F. Rex (Princeton: Princeton University Press, 1990).

[38]Claude E. Shannon and Warren Weaver, The Mathematical Theory of Communication (Urbana: University of Illinois Press, 1949).

[39]Warren Weaver offered this explanation in his essay interpreting Shannon's theory in Shannon and Weaver (1949).

[40]See N. Katherine Hayles, Chaos Bound: Orderly Disorder in Contemporary Literature and Science (Ithaca: Cornell University Press, 1990); for a major statement of this thesis, see Ilya Prigogine and Isabelle Stengers, Order Out of Chaos: Man's New Dialogue with Nature (New York: Bantam, 1984).

[41]Wiener's views on photosynthesis and Maxwell's Demon are discussed by Masani, pp. 155-56. See also Norbert Wiener, "Cybernetics (Light and Maxwell's Demon), Scientia (Italy) 87 (1952): 233-235; reprinted in Coll. Works IV, pp. 203-05.

[42]Michael Serres plays multiple riffs upon this idea in The Parasite, translated by Lawrence R. Schehr (Baltimore: Johns Hopkins University Press, 1982).

[43]Valentino Braitenberg delightfully explores the possibility that simple machines can demonstrate behavioral equivalents to emotional states, including fear, love, envy and ambition, by constructed a series of "thought machines" (machines designed in principle but not actually built). See Vehicles: Experiments in Synthetic Psychology (Cambridge: MIT Press, 1984).

[44]Galison, “Ontology of the Enemy,” p. 232.

[45]Despite Wiener's efforts, after World War II cybernetics became more, not less, entangled with military projects. The close connection between the military and cybernetics is detailed by Paul N. Edwards, The Closed World: Computers and the Politics of Discourse in Cold War America (Cambridge: MIT Press, 1996) and Cyborg Worlds: The Military Information Society, edited by Les Levidow and Kevin Robins (London: Free Association Books, 1989).

[46]Norbert Wiener, Cybernetics: Or, Control and Communication in the Animal and in the Machine (Cambridge: MIT Press, 1961, 2nd edition; 1st edition, 1948).

[47]Richard Dawkins, The Selfish Gene (New York: Oxford University Press, 1976).