"The scientist," he declared late in life,"is always working to discover the order and organization of the universe,and is thus playing a game against the arch enemy, disorganization. Isthis devil Manichean or Augustinian? Is it a contrary force opposed toorder or is it the very absence of order itself?"

Wienerqueries whether the world is an active (Manichean) opponent or merelya passive (Augustinian) antagonist, the only difference being that the"Manichean devil" used tricks, craftiness, and dissimulation against us,while the "Augustinian devil" did not change methods: "The differencebetween these two sorts of demons will make itself apparent in the tacticsto be used against them" (HU, p. 35).

eltan-schauung,

Differing from Hobbes (Wiener did not believethat people were fundamentally selfish), he nonetheless saw morality asa conflict not resolved in the distant past but as continuing into the hereand now.

reinstantiated

At the same time, it would clearly be erroneous toview cybernetics as a logically impelled set of beliefs

leatory

us, nor are we the model upon which the Other is empathetically formed;our understanding of the cybernetic Enemy Other becomes the basis onwhich we understand ourselves. Wiener's image of the human and natu-ral world is, in the end, a globalized, even metaphysical, extension of theepochal struggle between the implacable enemy from the sky and theAllies' calculating AA predictor that did battle from the ground. It is animage of human relations thoroughly grounded in the design and manu-facture of wartime servomechanisms and extended, in the ultimate gen-eralization, to a universe of black-box monads.

If this cybernetic conception seems to differ from more familiar con-ceptions of the Other, it should. The cybernetic Enemy Other has littleto do with the racialized Other so horrifyingly invoked by Blamey, andexamined, for example, by Edward Said in Oriental~sm.~'There is nosense in which Wiener sees the German bomber pilot as a racially lesserbeing. Nor is the German pilot an Other in being simply invisible. Finally,I take it to go without need of much elaboration that the servomechanicalpilot is not Emmanuel Levinas's Other, where the recognition of the in-eradicable humanity outside of oneself is the fundamental move in theestablishment of an ethical p h i l o s ~ p h y . ~ ~No, Wiener's conception of theEnemy Other is more like his depiction of the game players in von Neu-mann's theory: "perfectly intelligent, perfectly ruthless operators" (C, p.159).This is a theoretical representation in which information, statistics,and strategies are applied to moves and countermoves in a world of op-posing but fundamentally like forces.

I wouldargue, is indissolubly tied to their genealogy. To understand the specificcultural meaning of the cybernetic devices is necessarily to track themback to the wartime vision of the pilot-as-servomechanism. In the air-ground battle, it was a short step for Wiener and Bigelow to take thepilot-as-servomechanism directly over into the AA gunner-as-servomech-anism and thence to the operation of the heart and proprioceptive senses.From the body, it was us more generally-we humans-whose intentionscould be seen as none other than self-correcting black-boxed entities andfinally nature itself that came to be seen as a correlated and characteristicset of input and output signals.

After the bomb,as we have seen, Wiener's association of danger and moral weight withcybernetics grew even stronger. War gave the new cybernetic technolog-ies a role to play in the Manichean drama of the world. Mere governors,thermostats, and voltage regulators could not usher in a cybernetic age-weapons could.

In World War 11, the mechanizedsoldier faced his opponent as a machine, and machines manifested them-selves as people.

While prewar behaviorists might have cau-tioned against the ascription of internal states, war made it impossible;reading the hidden enemy meant reading his actions. In the mechanizedbattlefield, in those life-and-death confrontations with an enshrouded en-emy, the identity of intention and self-correction was sustainable, reason-able, even "obvious."

Tothe thousands of servicemen who used and faced this new generation ofweapons, the "human" character of self-regulating machines seemed alltoo human. After all, trying to shoot down a Junkers JU 88 heading forLondon or a V-1 buzz bomb doing the same thing was not all that differ-ent. A skipper trying to dodge a self-guided torpedo could be excusedfor referring to the device as "trying" to kill him, as could the pilot ascrib-ing airfoil self-adjustment to the work of "gremlins."

What we have seen in Wiener's cybernetics is the establishment of afield of meanings grounded not through zeitgeist but explicitly in theexperiences of war

collocation

Wiener had come to see the human-machine relation as a model, if notan incarnation of the bond between God and "man." The paradoxes ofreligion ("Can God create a rock too great for him to move?") reemergedas questions about the cyberneticist and his offspring ("Can a human cre-ate an entity that can beat him at chess?"). On the last lines of the lastpage of his last book, Wiener put it this way: "Since I have insisted upondiscussing creative activity under one heading, and in not parceling it outinto separate pieces belonging to God, to man, and to the machine, I donot consider that I have taken more than an author's normal liberty in

In choosing the cyborg to lead theflight from modernism, one risks reducing the picture of human capaci-ties to one of tactical moves and countermoves in a metaphorical exten-sion of automatic airwar.

We who make cyborgs are, in the end, like gods

t machines could be open-ended, nondedicatedin their function, and able to reproduce, learn, and interconnect with thehuman. But Wiener, unlike Haraway, saw power and control as absolutelycentral to the very definition of cybernetics, for better or worse

For the classic cyberneticists (exemplified by Wiener, Rosenblueth,McCulloch, and their colleagues), the blurred boundary between humanand machine opened an infinity of possibilities

After all, the very notion of a cyborg issuedfrom an Air Force contractor's extension of Wiener's ideasx2I would ar-gue that the associations of cybernetics (and the cyborg) with weapons,oppositional tactics, and the black-box conception of human nature donot so simply melt away

. ~ ~Haraway alludedto the "cyborg orgy" that she saw "coded by CSI,command-control-com-munication-intelligence, an $84 billion item in 1984's US defence bud-get."

In her view, cybernetics, al-though often used to sanction the status quo, is ultimately far more opento a new and more liberating vision of the biological sciences than thepsychobiological and organic functionalist theories that preceded it. Thecybernetic biological view (sociobiology) is, in Haraway's view, less opento racism or sexism because in cybernetics the organic body is depictedas an engineering entity, always modifiable, and never defined essen-tially.

hese included intelligence testing, human relations,physiology, and racial hygiene. After the war, the new sciences of informa-tion- and control-dominated systems reshaped biology, including sociobi-ology. This new, more cybernetic biology emphasized communicationand feedback.

we track Lyotard's postmod-ernist and game-theoretical worldview back deep into the heart of theManichean sciences. As we study the development of postwar science,then, it seems to me of utmost importance not to seize uncritically thecentral metaphors of operational analysis, game theory, and cyberneticsand make them our own while claiming all the while a new "postmodern"periodization

Formally, militarily, and philo-sophically, theirs was a universe of confrontation between opponents:Allies to Axis, monad to monad, message to message, and mechanized"man" to servomechanical enemy

"a cybernetic machine doesindeed run on information, but the goals programmed into it [leave noway] to correct in the course of its functioning . . . its own performance."Nothing could be further from the mark. This self-correction is exactlywhat Wiener's machines did.

Here at leasttwo of Lyotard's categories (denotative and prescriptive) directly parallelWiener's distinction between the indicative and imperative moods ofmessages

who were blurring the boundaries between the wet sci-ences of the brain, psychological properties, and the machine

We are truly, in this view of theworld, like black boxes with inputs and outputs and no access to our oranyone else's inner life

uasi-solipsism

At the sametime, Wiener wanted to make plain that while epistemology may wellcapture the knowledge-gathering function of the science, cybernetics willnot rest there: "messages may be sent for the purpose of exploring theuniverse, but they may also be sent with the intention of controlling theuniverse." Precisely because Wiener wanted to accentuate the dual aspectof information, he distinguished between messages that could be sent "inthe indicative and the imperative mood."

Within the rubric of "purposeful behavior," then, Wiener and hiscollaborators Bigelow and Rosenblueth allowed for those acts that do notinvolve feedback while the process is underway (such as a frog that shootsits tongue out towards a fly) and those (such as a self-guided missile ortorpedo) that gather information and use that information to correctthemselves en route. But beyond any particular features of humans ormachines lay Wiener's deep-seated commitment to a behaviorist visionof both. His was not a claim that no criteria differentiated humans andmachines. Quite obviously there was no machine that could (as yet) writea Sanskrit-Mandarin dictionary; and, similarly, no living organism rolledon wheels. But it was the behaviorist impulse to focus on broad classes ofactions, and to do so on the basis of the input and output he knew so wellfrom communication technology, that led Wiener to his blurring of theman-machine boundary. Black boxes, as Wiener used the term, meant aunit designed to perform a function before one knew how it functioned;white boxes designated that one also specified the inner mechanism

Explaining that they did not carewhether, in the abstract, machines "are or can be like men," Rosenbluethand Wiener insisted that the question was "irrelevant" for scientific objec-tives:

Whether or not we accept Wiener'stechno-periodization of the history of humanity, there seems little doubtthat he and many of his contemporaries saw themselves as standing at ahistorical and philosophical watershed in which the Manichean scienceswould undergird the cybernetic age

Paradoxically, during the warWiener had extended the cybernetic vision beyond its narrow applica-tions because of the weakness of the AA predictor; now that he associatedcybernetics with the power of cataclysmic weapons, he tried to push cy-bernetics away from the military arena because of its deadly efficacy. Ei-ther way, for Wiener and many colleagues, the association of cyberneticswith its wartime origin was forcefully and deeply inscribed in the culturalmeaning of the new science and its machines.

In the years that followed, Wiener repeatedly stressed the power of cy-bernetics to save, enslave, or destroy humanity.

But, agonizing over the possible uses ofhis research, he halted his letter with these "sketchy" descriptions in partbecause of a claimed modesty and in part because "in these troubledtimes I do not feel any too certain that I shall continue in science indefi-nitely. I do not know how to publish work without making it available forthe strongest hands and I do not like the strongest hands of the presenttime. I feel it most intensely personally and in particular what I haveseen in looking upon a world completely inadequate to receive the atomicbomb."

Ever since the atomic bomb fell I have been recovering from an acuteattack of conscience as one of the scientists who has been doing warwork and who has seen his war work a[s] part of a larger body whichis being used in a way of which I do not approve and over which Ihave absolutely no control. I think the omens for a third world warare black and I have no intention of letting my services be used insuch a conflict. I have seriously considered the possibility of givingup my scientific productive effort because I know no way to publishwithout letting my inventions go to the wrong hands.

The model then expanded to become anew science known after the war as "cybernetics," a science that wouldembrace intentionality, learning, and much else within the human mind.Finally, the AA predictor, along with its associated engineering notions offeedback systems and black boxes, became, for Wiener, the model for a cy-bernetic understanding of the universe itself. This paper is an explorationof that expansion.

But Wiener's elec-tronic manipulation did not stop with halting Nazi air attacks. In thecourse of characterizing the enemy pilot's actions and designing a ma-chine to forecast his future moves, Wiener's ambitions rose beyond thepilot, even beyond the World War. Step by step, Wiener came to see thepredictor as a prototype not only of the mind of an inaccessible Axis op-ponent but of the Allied antiaircraft gunner as well, and then even morewidely to include the vast array of human proprioceptive and electro-physiological feedback systems

proprioceptive

otalization

Finally, for Wiener, the present age, usheredin by the vast array of electromechanical devices of the war, was the ageof information and control. If these developments reached back to Kelvinand Gauss, they found their real form (and interpreters) only in the labo-ratories and factories of radar and its associated systems. This age, ourage, was that of the servomechanism

To an antiaircraft operator, the enemy really does actlike an autocorrelated servomechanism.

aperp

If humans do not differ from machines from the "scientific stand-point," it is because the scientific standpoint of the 1940s was one of men-machines at war.

Taylor who could not abidethe elimination of inner states of human intention, desire, pleasure, andpain in favor of purely observable manifestations.

When I consider how the psychologists have been trying to solveexactly this problem for decades (if not for centuries), the black boxbeing the brain, and when I think how little attention they havegiven to the principles involved, my opinion of psychologists falls toa new low. The trouble with the psychologist is that he is too proudto learn to walk before he tries to run. So today he lies on his back,foolishly waving his legs, and pretending to be a ballet-dancer, whenin reality he hasn't yet learned how to crawl.For this reason I regardit as highly complimentary when I say that your study of the "blackbox" problem is a first step towards a scientific psy~holog

potheosis

anegyrics

"That's about what a person is, re all^.

ntention, like desire, is something that is justas real as more tangible acts. Taylor protested: We should not abandonconcepts simply because they are not operationally useful to ~cienc

We should read this last remark critically and historically. In 1941 and1942, it had made sense to Wiener and his collaborators to view humans,qua pilots and gunners, as undifferentiated from the bombers and anti-aircraft units in which they fought. Seen as man-machine enemies, from amilitary perspective "humans do not differ from machines." It was thena short step from viewing the enemy as a cybernetic entity to seeing thequasi-automated Allied aircraft gunner the same way. What had begunin the Manichean field of science-assisted warfare had now been decon-textualized. By 1950, Wiener had globalized his claim: under the gazeof scientific inquiry, human intentionality did not differ from the self-regulation of machines, full stop

Because many of the relevant developments were directly tied to the wareffort, Wiener asserted, the assembly would necessarily be nonpublic. Itwas a new vision of the world that was to emerge from this secret conflu-ence of war sciences, one that would embrace matters of "engineering,physical, and even economic and social interest."49 Wiener, Aiken, andvon Neumann named the group the "Teleological Society."

If I could not find these talents joined together in a single man, Iwould be forced to assemble a team of people each with particulartalents in one field and a general knowledge of the others. In thisteam I would probably be the only mathematician thus the projectas a whole would concern other engineering groups as much as theAmerican Mathematical Society and it would be necessary for me tocross ordinary professional lines. l

Behaviorists took note. The eminent Harvard psychologist and his-torian of psychology Edwin Boring found Wiener's suggestion that allfunctions of the brain might be duplicated by electrical systems "very at-tractive." Having had a chance to contemplate this circuit-reductionistprogram, Boring reckoned in a 13 November 1944 letter to Wiener thathe could provide "a pretty complete list of psychological functions" in hisspare time, all psychological facts being in principle expressible in termsof stimulus and response. "A symbolic process" would be "a delayed, ade-quately differential reaction"; "introspection" would be a reaction to areaction, and Wiener's task, should he decide to accept Boring's chal-lenge, would be to transfer these stimulus-response pairs and respondwith his own matching catalogue that would give the "same specificity of'output' to 'input."' With fourteen psychological properties on the listalready and others like "Generalization" and "Abstraction" to be added,Boring assured Wiener that a paper with these electrical designs wouldgreatly benefit "us operationally-minded psychologists." "Is it a go?" Bor-ing queried. "I do not know that you can [do it], but I should be bettingon Black-box engineering now had something more complexthan electrical amplification as its functional goal: to re-create the minditself

But he and Rosen-blueth (who collaborated on the riposte) reemphasized that the weightedwheel and the magnetic compass differ from the servomechanisms ofguided missiles and AA predictors because the former are passivewhereas the latter are active. In the laboratory, Wiener and Rosenbluethinsisted, the physics and engineering practices behind self-regulating sys-tems are utterly different from that of bricks and clocks; the former aregoverned by time-reversible causal stories whereas the latter are unidi-rectional in time. (The AA predictor, for example, makes its statisticalforecast on the basis of the histop of the pilot's past performance.)

uld be rendered into a purposeful machine by theaddition of a lead weight on its perimeter? Even the guided missile, thatparagon of purpose, is hardly to be philosophically distinguished fromnon-self-regulating devices. Consider a missile following mechanicallyalong a taut wire attached to a target. Such a mechanism might be lessetherial than radar guidance but would hardly be distinct insofar as itcould be considered self-regulating. "The expression 'target-seeking mis-sile' is," Taylor concluded, "metaphorical." Wiener and Bigelow mightchoose to redefine the very concept of purpose, but their discovery wouldamount to no more than the redefinition of the plus sign with that ofmultiplication: "entirely correct, but scarcely significant."

etherial

Thiswas fine with Wiener, though he found von Neumann's sketch to lack thecrucial transition from the computing machine to the control machine."The issues that come up here are those of transfer from continued datato counted data; of the final transition from counted data to the motionof a shaft effector; and the sensing of the motion of the effector by feed-back or other quasi-proprioceptor apparatus." Such a feedback system,which Wiener had stressed from his earliest work on servomechanisms,continued to occupy a central place in his thinking. For it was this sameproprioceptive process that occurred in mechanical controls, organiccontrols, and in hybrid mechanico-organic systems. Von Neumann cededthe point.j4Despite the fantastic array of supporters that Wiener's approach elic-ited, there was resistance. In 1950 Richard Taylor, a young philosopherfrom Brown University, asked with incredulity if Wiener and his collabo-rators could seriously be proposing a definition of purposefulness thatwas built purely on the culmination of a sequence of events. Rosenblueth,Wiener, and Bigelow's definition was this:The term purposeful is meant to denote that the act or behavior maybe interpreted as directed to the attainment of a goal-i.e., to a finalcondition in which the behaving object reaches a definite correlationin time or in space with respect to another object or event. Purpose-less behavior then is that which is not interpreted as directed to agoal.j3To Taylor this definition was both so all-encompassing as to rule out noth-ing and so devoid of content that it had no overlap with any commonmeaning of the term. Let a clock run for many years only to break downat midnight on New Year's Eve. What rules this out as an instance ofpurposefulness? A brick tumbles off a building, killing a passer-by. Is this,too, to be considered as purposeful? For Taylor, the utter arbitrarinessimposed by the clause "'may be interpreted"' makes it all too easy to allowthese tumbling bricks and failing clocks to be counted as purposeful,voiding the term of any similarity with our usual ~ n d e r s t a n d i n g .

Within a few weeks, Wiener's ambition left behind even the humanmind.

Our consciousness of will in another person, Wiener argued, is just thatsense of encountering a self-maintaining mechanism aiding or opposingour actions. By providing such a self-stabilizing resistance, the airplaneacts as if it had purpose, in short, as if it were inhabited by a gremlin

The semi-humorous superstition of the gremlin among the aviatorswas probably due, as much as anything else, to the habit of dealingwith a machine with a large number of built-in feedbacks whichmight be interpreted as friendly or hostile. For example the wings ofan airplane are deliberately built in such a manner as to stabilize theplane, and this stabilization, which is of the nature of a feedback . . .may easily be felt as a personality to be antagonized when the planeis forced into unusual maneuvers.-

Darwin's dog suffered remorse; Wiener's AA predictor had fore-sight.

What Wiener was willing to do, even in the worst days of war, was toturn to psychological and philosophical implications of the predictor.In their 1943 article "Behavior, Purpose and Teleology," Wiener andBigelow collaborated with the cardiologist Arturo Rosenblueth, then vis-iting Harvard Medical School, to present a new, behaviorist descriptionof the very concept of purpose. Aside from the pure satisfaction of classi-fication, the authors were pleased to single out the class of predictivebehavior because "it suggests the possibility of systematizing increasinglymore complex tests of the behavior of o r g a n i ~ r n s . " ~ ~Of particular impor-tance, they contended that their classification rehabilitated "purpose"and "teleology" by bringing them under the aegis of a "uniform behavior-istic analysis" that was equally applicable to living organisms and ma-chines.

"teleology"

judged the only hope for the method to lie in a vastly increased statisticalbase involving the calculation of tens, if not hundreds, of tracks. Sincethis would tie up the computing facilities of the country, and because thelikelihood of improvement struck him as "too distant to be significant inthe present war," Wiener hesitated to recommend further research untilafter the end of the war.41What went wrong? Wiener speculated:To what extent the negative result of this investigation is due to badtracking, to what extent to the restriction of the useable past [flightpath] to 10 seconds, and to what extent to the fact that the enemyplane has a very considerable chance to change its flight pattern,whether voluntary or involuntary, in the twenty seconds of projectileflight, is not yet fully clear.42It may have been "not yet fully clear," but Wiener was "convinced" thatit was the enemy's capacity to maneuver rather than anything else thatwould save him from inevitable destruction at the mechanical hands ofthe predictor. Failure came hard, for Wiener was frustrated by the pre-dictor's weakness: "I still wish that I had been able to produce somethingto kill a few of the enemy instead merely of showing how not to try tokill them."

Even in the midst of a war project that did not yet approach fieldcapability, Wiener clearly had already begun to reflect on the broaderramifications of this species of machine. On the same day he saw the pre-dictor demonstrated, 1 July 1942, Stibitz recorded Wiener's wider ambi-tion for the device:W[iener] points out that their equipment is probably one of the clos-est mechanical approaches ever made to physiological behavior. Par-enthetically, the Wiener predictor is based on good behavioristicideas, since it tries to predict the future actions of an organism notby studying the structure of the organism but by studying the pastbehavior of the organism

Most of the day is spent with Wiener, Bigelow, and Mooney. It simplymust be agreed that, taking into account the character of the inputdata, their statistical predictor accomplishes miracles. Whether thisis a useful miracle or a useless miracle, W[arren] W[eaver] is not yetconvinced. The fact that predictions can at present be made only fora maximum of 2 seconds is a very serious limitation. . . . For a1-second lead the behavior of their instrument is positively un-canny.37WW threatens to bring along a hack saw on the next visitand cut through the legs of the table to see if they do not have somehidden wires somewhere

W h e t h e rhis remarks were a sponta-neous expression of excitement over the new results or a crypticdeclaration of a priority claim, Wiener clearly saw the AA predictor, evenbefore it was ready to shoot down a plane, as the prototype of a newbehaviorist understanding of the nervous system itself. By the time Wie-ner wrote Haldane, he was in the final stages of preparing the machinefor its great unveiling.

Unmentioned was the content of these behaviorist studies. For securityreasons Wiener would not reveal that the time-pattern behaviors werethe pilot's evasive maneuvers and the test procedures Wiener employedto reproduce these patterns from the responses of test-subjects in thesafety of the 1aborato1-

Behaviorism as we all know is an established method of biologicaland psychological study but I have nowhere seen an adequate at-tempt to analyze the intrinsic possibilities of types of behavior. Thishas become necessary to me in connection with the design of appara-tus to accomplish specific purposes in the way of the repetition andmodification of time pattern^.^

Weaver replied the next day, after seeing Wiener pacing furiously up anddown a room, "perspiring profusely," and apologizing for being unableto transform an integral into a more easily calculable, rapidly convergingseries that the great statistician Jerzy Neyman could use. "Upon inquiry,"Weaver concluded, "it turned out that [Wiener] had not been doing anyof the things we particularly wanted him to d o and that his busyness con-sisted of 'holding myself in readiness in case other jobs turned up."'

He seems in an unusually bad nervous state the last few days, and Ihave been trying to get him to take a few days' rest. He had an unfor-tunate clash with the cleared patent attorney whom M.I.T. had askedto study some of his ideas on circuit theory, and at the same timehe felt that the Radiation Laboratory was unappreciative of certainsuggestions he had made to them on filter design. As a result of hisstate, Bigelow seems somewhat distracted, but I hope before verylong this part of the zoo will be quiet agaim

What the "big shots" lacked, Wiener contended, was a deep understand-ing of, for example, Brownian motion and generalized harmonic analy-sis. These were areas that Wiener had contributed to in fundamentalways; having found himself shunted off to trivial problem solving, he wasfurious. Without the requisite communications knowledge, Wienerprophesied, "the military efforts of the Laboratory will be about at a goodboy-scout level."g

Pressure and frustration began to overcome Wiener. He was workingfrantically, often with the powerful stimulation of B e n ~ e d r i n e

and to turn such an individual loose in your laboratory without spe-cial training, no matter what a big shot he may be in his own subject,is like ordering a corn-doctor to amputate a leg. Better three weeksdelay while the big shot is learning his new trade than three monthsof puerilities and blunders.

puerilitie

It does not seem even remotely possible to eliminate the human ele-ment as far as it shows itself in enemy behavior. Therefore, in orderto obtain as complete a mathematical treatment as possible of theover-all control problem, it is necessary to assimilate the differentparts of the system to a single basis, either human or mechanical.Since our understanding of the mechanical elements of gun pointingappeared to us to be far ahead of our psychological understanding,we chose to try to find a mechanical analogue of the gun pointer andthe airplane pilot.

Many other extensions of these ideas will suggest thenlselves to the physi-ologist, the neuropathologist, and the expert in aptitude test^."'^ More tothe point, it suggested that a more refined AA predictor would use apilot's own characteristic flight patterns to calculate his particular futuremoves and to kill him.

In particular, the pseudo pilots' "nervous reactions" exhibited twocrucial features. First, there was no particular correlation among the re-corded fluctuations of different operators; second, there was a high de-gree of autocorrelation between an operator's earlier and subsequentperformances. More specifically, Wiener chose the following definition ofprediction. Imagine a number of flight paths (ten, for example) that allcoincide for a given segment of their trajectory but may differ after agiven time, t. Now pick a point in space where we expect a plane to beat, say, t + 2 seconds. For any such predicted point we can calculate thesquare of the difference between the predicted point and the actual posi-tion of the first plane at t + 2 seconds, and we can do the same for theother nine planes. The point for which the sum of squared errors is mini-mized is what Wiener calls the best predi~tion.'~It turned out that predic-tion worked rather badly for one operator based on another operator'sdata, but any given operator was enormously self-consistent.

Quite deliberately, the experimenters made the task exceed-ingly difficult by racing the target across the wall at high speed and byinserting a mechanical lag into the control stick. This, they hoped, wouldcreate precisely the disassociation between kinaesthetic sense and visualinformation that the pilot had to face in the theater of war and wouldtherefore lead to the same sort of feedback difficulties

Here was a problem simultaneously physical and physiological: the pilot,flying amidst the explosion of flak, the turbulence of air, and the sweepof searchlights, trying to guide an airplane to a target. As Wiener saw it,humans acting under stress tend to perform repetitively and thereforepredictably.

We realized that the "randomness" or irregularity of an airplane'spath is introduced by the pilot; that in attempting to force his dy-namic craft to execute a useful manoeuver, such as straight-line flightor 180 degree turn, the pilot behaves like a servo-mechanism, attemptingto overcome the intrinsic lag due to the dynamics of his plane as aphysical system, in response to a stimulus which increases in intensitywith the degree to which he has failed to accomplish his task. A fur-ther factor of importance was that the pilot's kinaesthetic reaction tothe motion of the plane is quite different from that which his othersenses would normally lead him to expect, so that for precision fly-ing, he must disassociate his kinaesthetic from his visual sense.'O

As the AA predictor came to fruition, Wienercame to see it as the articulated prototype for a new understanding of thehuman-machine relation, one that made soldier, calculator, and fire-power into a single integrated system. His two thousand-odd dollarswould be conceptually stretched to blanket the earth

To be useful to the war effort, it was sci-ence itself that would have to change, becoming both materially groundedand squarely directed into the world of weapons. For Wiener, Morse, andtheir colleagues, science was at war, even if the country was not.

No one who did not already have a feeling for engineering, at least bytaking on the construction of radio sets as a hobby, could even hope toparticipate. "There is nothing in a drawing in abstract algebra or topol-ogy . . . which would prepare one in any way to cooperate in engineeringdesign."

Wiener used Bush'spartial differential equation machine as an example ofjust how far math-ematics would have to go if it were to advance. It could not remain "ex-clusively in the hands of mathematicians." He would need someonealready versed in computing technique, someone, say, from Remington-Rand or IBM. This man (his term) would have to be familiar with vacuumtube work, and he would without question have to know his way aroundcommunication engineering techniques such as telephone technology.

Partly as a result of an extended exchange and collaboration withBush and partly as a result of his own early circuit designs, Wiener overthe next few months came to insist that the wartime mathematician hadto stray far from the "purity" of the prewar years.

It was this spate ofNazi bombing that precipitated Wiener's primary war work. By 23 July,Wiener had received notice that the armed forces had in hand his sugges-tion about the use of incendiary bombs and his reiterated desire to partic-ipate in the war effort.I3 But the main line of Wiener's military workcentered on a general theoretical and practical inquiry into the possibilityof radically improving antiaircraft technology. At root, Wiener's idea wasto use electrical networks to determine, several seconds in advance,where an attacking plane would be and to use that knowledge to directartillery fire. In an early simulated run of the AA predictor (November1940), Wiener wanted to see how his machine would fare in four cases: astraight-line bomber trajectory, another having twice the slope of the first,a third with a parabolic slope, and a fourth with a curve following theintegrated area of a semicircle. Since the circuit itself did not exist in wireand tubes, Wiener concocted a virtual mechanical equivalent and "tested"all four cases in simulated form on Vannevar Bush's differential ana-lyzer.

Already in February of 1940, he joined a subcommitteeunder the direction of the Princeton mathematician Marston Morse toconsider how the American Mathematical Society might contribute to anational emergency "which we hope will never arise."12 Needless to say,it did. Beginning on 10 July, German bombing attacks on British convoysgrew significantly in number, ushering in the Battle of Britain.

Norbert Wiener's upbringing resembled none so much as John Stu-art Mill's. His father, Leo Wiener, was an erudite and driven HarvardSlavicist who was determined that his son know languages, mathematics,and the sciences long before he was old enough to attend grade school.Graduated from Tufts University at the age of fourteen, and armed witha Harvard Ph.D. at eighteen, the younger Wiener found his own style ofwork in physics-based mathematics. By the late 1930s, Wiener was in hisforties and a major figure at MIT, where he had contributed to quantumtheory, ballistics, the theory of integration, and communication technol-ogy.

erudit

Cybernetics no longer appears as a futuristic bandwagon or as a ris-ing worldview that will leave mere mechanism in the dustbin of history,but it has much to tell us about the nature of the sciences in the mid-twentieth century and, as I will speculate, about postmodern theory inthe late twentieth century.

It was a vision in which the enemy pilotwas so merged with machinery that (his) human-nonhuman status wasblurred. In fighting this cybernetic enemy, Wiener and his team beganto conceive of the Allied antiaircraft operators as resembling the foe, andit was a short step from this elision of the human and the nonhuman inthe ally to a blurring of the human-machine boundary in general. Theservomechanical enemy became, in the cybernetic vision of the 1940s, theprototype for human physiology and, ultimately, for all of human nature.Then, in a final move of totalization, Wiener vaulted cybernetics to a phi-losophy of nature, in which nature itself became an unknowable but pas-sive opponent-the Augustinian devil.

I will argue that the system of weaponry and people that Wiener hadin mind was predicated on a picture of a particular kind of enemy. Onthe mechanized battlefield, the enemy was neither invisible nor irra-tional; this was an enemy at home in the world of strategy, tactics, andmaneuver, all the while thoroughly inaccessible to us, separated by a gulfof distance, speed, and metal.

Working with the Greek word for steersman, Wiener coined the termcybernetics in the summer of 1947 to designate what he hoped would be anew science of control mechanisms in which the exchange of informationwould play a central role.I0 If antisubmarine warfare was the formativeproblem for operations research, antiaircraft fire control was the key tocybernetics." Faced with the problem of hitting fast maneuverable bomb-ers with ground-based artillery, Wiener brought to bear his own estab-lished interest in feedback mechanisms, communication technology, andnonlinear processes

game theory postulated a logical but cunning opponent; it was designedprecisely to analyze an antagonist who played against us and would bluffto win

On the Allied side, three closely related sciences engaged this calcu-lating Enemy Other: operations research, game theory, and cybernetics.Each had its own prototypical war problem. Operations research focused,for example, on maximizing efficiency in locating and destroyingGerman U-boats in the North Atlantic and along the coast of the Arneri-cas.Wame theory, though it had mathematical roots in the interwaryears, exploded into view with John von Neumann and Oscar Morgen-stern's masterwork of 1944, Theory of Games and Economic B e h a v i ~ r ; ~strate-gists picked up the technique as a way of analyzing what two opposingforces ought to do when each expected the other to act in a maximallyrational way but were ignorant both of the opponent's specific intentionsand of the enemy's choice of where to bluff. Wiener, the spokesman andadvocate of cybernetics, in a distinction of great importance to him, di-vided the devils facing us in two sorts. One was the "Manichean devil""who is determined on victory and will use any trick of craftiness or dis-simulation to obtain this victory." Wiener's rational Manichean devilcould, for example, change strategy to outwit us. By contrast, the other,the "Augustinian devil" (and Wiener counted the forces of nature as such)was characterized by the "evil" of chance and disorder but could notchange the rules.'

After openinga spate of airmen's letters, one British censor from the Air Ministry re-ported on 21 June 1942: "[The letters] illustrate the effect of airmen'sremoteness from their attacks on human beings. Expressions of satisfac-tion that the Germans are having to undergo the punishment they havehitherto meted out to others are found in almost all letters, but there isan absence of vindictiveness or fanaticism in the phrases used.

This enemy's humanity was compromised not by beingsubhuman, vicious, abnormal, or primitive but by occupying physical andmoral distance.

"fighting Japs is not like fightingnormal human beings. . . . The Jap is a little barbarian. . . . We are notdealing with humans as we know them. We are dealing with somethingprimitive. Our troops have the right view of the Japs. They regard themas ~ e r m i n . " : ~These monstrous, racialized images of hate certainly pre-sented one version of the World War I1 enemy, but it was by no meansthe only one.

Enemies were not all alike. In the killing frenzy of World War 11, oneversion of the Enemy Other (not Wiener's) was barely human; to theAmericans, British, and Australians, the Japanese soldiers were oftenthought of as lice, ants, or vermin to be eradicated.

Over the next few years, Wiener's attention focused increasingly onthe problem of destroying enemy airplanes. His early efforts at computa-tion and antiaircraft fire coalesced in a remarkably ambitious calculatingdevice that he called the "antiaircraft (A4)predictor," designed to charac-terize an enemy pilot's zigzagging flight, anticipate his future position,and launch an antiaircraft shell to down his plane.

For the German AirForce had dubbed 13 August 1940 "The Day of the Eagle," and with itthe Battle of Britain had begun with an assault of almost 1500 aircraftflown against British air stations and aircraft factories. During the follow-ing two weeks over a thousand Londoners had died under the rain ofbombs, and September was worse. On 7 September alone, 448 civiliansperished; on 15 September the Germans pitched 230 bombers and 700fighters against London, Southampton, Bristol, Cardiff, Liverpool, andManchester.

He suggested proce-dures to improve Bush's computational device, the so-called differentialanalyzer, in ways that would facilitate faster design of war materiel fromairplane wings to ballistic shells. More concretely, he reiterated a previousproposal that the Allies loft air-bursting containers of liquified ethylene,propane, or acetylene gases to engulf a wide volume of the sky in a pro-

"I . . . hope you can find some corner of activity in which I may be of useduring the emergency," the mathematician and physicist Norbert Wienerwrote the czar of American war research, Vannevar Bush, on 20 Septem-ber 1940