Kell Robinson

Agency is an anxious topic and many of the forms of anxiety surrounding it in our own time can be traced back to the later nineteenth century, when electrical technologies began to link bodies and machines in continuous circuits of activity. What David Seltzer has called "melodramas of uncertain agency" arose from this circuitry as "the machine likeness of persons and the personation of machines" evoked a strange mixture of politics, science, and sorcery.

The mixture was at its most potent where machine circuitry was at its most intricate: in automata and calculating machines, whose behaviorimitated human coordination in ways that suggested the presence ofintelligence and volition. As the automatic machine became increasinglysuggestive of agency, any appearance of the automatic in human behaviorconversely seemed to suggest loss of agency. It was as though agencycould leak from bodies into machines through the circuitry by which they were interconnected.

James R. Beniger identifies "a crucial transition in human thought about programming and control" between the later nineteenth century and the 1930s, which he identifies as "the first stage of the revolution in control technology."

This transition in thought is accompanied by intense activity in the cultural imaginary, as anxieties about programming and control are thematised in painting, fiction, film, and dramatic literature. The performer embodies the human-machine confusion thematised by writers and scenographers and can therefore be said to engage with it more immediately. In performance, automatism is associated with enchanted beings (swan maidens, animated dolls), with puppetry and with "wooden" actors who can express nothing more than careful programming by their trainers. Yet by playing across the borderline between the agentic and the automatic, the performer can explore some of the disturbing ambiguities associated with the machine's uncanny lack of agency. Perhaps the performer and the machine have some strange affinity that draws out cultural anxieties about becoming automatic. When it comes to dealing with live presence--simulating it, or conjuring it forth, and setting it in dynamic relation with other presences--there is also a "live" concernw with agency. The electricity is on, moving through bodies, and the spectator can be witness to processes of energy transmission and conversion.

What I am calling "transferred agency" is a classic example of a cultural anxiety. Although we have well-established methods for addressing anxiety at the level of the individual, relatively little has been done to establish ways of addressing the cultural production of anxiety. Cultural anxieties are generated through narratives, tropes, and thematic paradigms. The anxious consciousness, especially when it is collective, is characteristically convinced that it is born of the present, its nerves attuned to a future stirring in the conditions of the moment. This in itself provides strong reason for studying its history. The cultural production of anxiety is obstinately formulaic and the strongest patterns of anxious speculation change very little from one era to the next, showing remarkable resistance to paradigm shifts occurring in fields of knowledge. Late-nineteenth-century anxieties about the transferral of agency from human to machine are repeated through the twentieth century, going beyond the 1930s into the era of the electronic revolution, adapting to changes in the nature and function of the machine without making changes to their own underlying logic. What follows is an attempt to observe the echoes in that underlying logic across four very differently placed versions of human/machine relations. Three of these have in common a live performative element that threatens to disrupt the substratum of logic and demonstrates how the evidence available from the bodies on view may serve at once to lure and to call the bluff on rhetorically formulated anxieties. The first account, though, may serve as a model against whichto measure the other three, as it sets out the logic in its purest form,in an anecdote that cuts a swath across the timeband by bringing together Charles Babbage (1792-1871), whose "Analytical Engine" is widely regarded as the forerunner of the electronic computer, and AlanTuring (1912-54), whose Turing test is still cited as the best technique for distinguishing human intelligence from that of the machine.

At the end of Godel, Escher, Bach, Douglas Hofstadter sets up a dialogue scene in which he as Author comes together with some of his characters to improvise music. The instruments surrounding them include an assortment of computers or, as the characters decide to call them, "smart-stupids." Suddenly, one of the smart-stupids announces the arrival of an unexpected guest. It is Charles Babbage. As the scene warms up, Babbage decides to perform a little theatrical gesture. He proposes, right there, before the eyes of his fellow improvisers, with nothing in his hand and nothing up his sleeve, to conjure into being an artificial intelligence:

Now, if I have not made too many errors, this smart-stupid will simulate a human being whose intelligence is six times greater than my own, and whom I have chosen to call "Alan Turing." This Turing will therefore be--oh, dare I be so bold as to say this myself?--moderately intelligent.

And sure enough, in a couple of beats, a figure appears on the screen. It is Alan Turing, and it says: If I have not made too many errors, this smart-stupid will simulate a human being whose intelligence is six times greater than my own, and whom I have chosen to call "Charles Babbage." This Babbage will therefore be--oh, dare I be so bold as to say this myself?--moderately intelligent. Anticipating trouble, one of the witnesses instantly tries to arbitrate: You, Alan Turing, are in the smart-stupid, and Charles Babbage has just programmed you! We just saw you being brought to life, moments ago. And we know that every statement you make to us is merely that of anautomaton: an unconscious, forced response.

But there is no way it's going to be as easy as that, and the bickering starts in earnest.

--I hope, gentlemen, that you'll forgive the rather impudent flavour of the preceding remark by the Turing Machine; Turing has turned out to be a little more belligerent and argumentative than I'd expected. --I hope, gentlemen, that you'll forgive the rather impudent flavour of the preceding remark by the Babbage Engine; Babbage has turned out to be a little more belligerent and argumentative than I'd expected --Being merely a program written by me, he harbours the illusion of having invented it all on his own!

--Me, a program written by you? I insist, sir, that matters are quite the other way round. Eventually, they are sent off to take the Turing test, as a tie breaker. The results are inconclusive and when they return, it is, of course, Babbage who is on the screen and Turing who is in real space. This is not a case of live performance. It is an elegant equation unsullied by the mess of differential detail that makes up embodied identities. The mathematical equivalence of Babbage and Turing depends upon their being abstracted from any kind of social formation. To place between these antagonists a line separating them in race, class, gender, or generation--or even one of physical typology (fat/thin, short/tall, sloppy/smart)--would give politically loaded meaning to their competition for agency. The logics here are algebraic and economic.There are ways in which creating a character for the stage and creating a technological model of the human are analogous. In particular, theyare attended by similar anxiety driven fantasies. What if the simulationof life turns into life itself? To create the illusion of autonomous being is to play with fire, like Prometheus; to dabble in the uncanny, like Hoffman's Dr. Coppelia. To push the limits of one's own agency by creating a being without agency is a dangerous move in terms of the psychical economy. How can the stock of agency remain all in one account, and why should it? Surely a being that is empty of agency must draw it from somewhere, and the only source to which it is connected is its own creator, who after all, deserves what is coming to him because, not content with making objects that are agency neutral, he has created an agency vacuum that must--automatically, so to speak--seek to fill itself.

In late-nineteenth-century thinking about the body, a concern toidentify how bodily energy behaves becomes bound up with attempts todefine agency. William James devoted persistent attention to the relationship between agency and energy in voluntary movement. In anessay of 1890, he offers a critique of "The Automaton Theory," as it is summarily expressed in T. H. Huxley's proposition that "the feeling of volition is not the cause of a voluntary act, but the symbol of thatstate of the brain which is the immediate cause of that act. We are conscious automata."

James's own view is that to "turn consciousness out of the door" is "an unwarrantable impertinence in the present state of psychology." Asserting that consciousness has a crucial role in the selection ofaction--in giving the go-ahead, or imposing a veto--he proposes a way ofdefining voluntary (as distinct from automatic) movement as expressive of a fiat delivered by the consciousness.

Energy and agency are here being read as signs of each other, and come to be used almost interchangeably by thinkers less discriminating than James. So when the first law of thermodynamics establishes that energy can be neither created nor annihilated, but only converted or transmitted, this is ghosted by an analogous proposition about agency.If energy is always seeking opportunities to convert or transmit itself, agency may, analogously, be in search of opportunities of transference or transformation.

The Rarefied Ornament

When the machine takes on the status of performer, it undergoes a transition from the functional to the marvellous. In its creation, the roles of the sorcerer and the technologist overlap. Automata as performing machines go back into ancient history and are associated with great centers of power: the courts of the Pharoes, of Chinese and Byzantine Emperors, of Russian Tsars. As a court performer, theautomaton could upstage the dwarfs, jugglers, and dancing slave girlswith its infinitely repetitious patterns of minute sound and movement.As an ambassadorial gift, it was also a strategic form of power play:our ruler sends to yours this intimation of the uncanny, an intricateswitching point between mechanism and vitalism, of which mystery only we know the secret. Don't mess with us. In May 1870, as the close of the second Empire was imminent, the Paris Opera staged Coppelia, the first example of a distinctively modern reversal in which the performer appears as automaton. This became a genre, with The Nutcracker (St. Petersburg, 1892) and Petrouchka (Paris, 1911) to follow as its two most celebrated examples. All three of these ballets give a central role to the manufacturer-magician whose creations threaten to escape his power as their movement in performance changes from mere activity into expressiveness, and so shows them "coming to life." This overseer is a figure belonging to an older genre. The sorcerer of the romantic ballet is an agency vampire whose power works all one way, turning his human victims into creatures whose loss of selfhood manifests as loss of the capacity for voluntary movement. Responding to an old dream of imperial power--in which its own concentrated volition reaches the point of invincibility while its agency-free subjects are becoming animal, becoming machine--this fantasy has its most compelling expression in Swan Lake. Swan Lake is a ballet blanc and a winter ballet; its formal premiere before the Tsar took place in St. Petersburg in January 1895. In the Petipa-Ivanov choreography (reworked from an earlier version of the ballet), the enchanted swan maidens behave suspiciously like automata, and nowhere more so than in the choreographic show-piece performed by the four cygnets, with arms linked and perfectly turned legs working in geometric formation. Are these the precursors of the Tiller Girls, demonstrating the literal connotations of the Royal Command Performance?

Meanwhile, working to cross the other way over the vital dividing line, the machinic performer was widening its sphere. As it entered its Golden Age, the later nineteenth century, the automaton appeared on the world stage at the Universal Exhibitions of 1867, 1878, and 1889 in Paris and discovered a global audience. More preciously intricate than ever before, and with a price tag to match, it was nevertheless developing a kleptomaniac taste for performance models stolen from the circus, the street and the music hall. Popular performers cannily took to reversing the tactics, developing a kind of cross-over game. Pierrot, as the most popular pantomime figure on the boulevards, became a set-piece in the major houses of automata. Vichy's Pierrot, one of the finest of these creations, was in turn adopted as a model for an act in the Ziegfeld Follies in which Bessie McCoy Davis was lowered from the flies perched on a crescent moon designed by Joseph Urban. McCoy Davis also imitated the costume and movements of an automaton for her performance as the "dancing Somnambulist" in the 1918 Follies presentation, Nine O'Clock Frolic. Anna Held, an early Ziegfeld star, played an animated doll in La Poupe (1897), modelling her gestures on those of the more elegant Parisian automata.

Banjo-playing black minstrels, whose images were freely plagiarised by the automaton houses, responded by stylising their movements into machinic jerkiness and wearing expressions of fixed exuberance on their faces. Eventually, it becomes impossible to tell from which side of the fence the vogue for satin clown costumes and heels-down-toes-up choreography has been appropriated. What is going on may be more in the nature of complicity than competition. The automaton is a reification of the performer, and when performers parody their own reification, the dynamics start to change. Questions arise over who is calling the tune,in contravention of a long history of performing to command. To besuccessful in vaudeville was to be allowed a large measure of caprice.Category games were played, in which cross dressing, blacking up,whiting out, or going ape echoed all the exotic games of the automaton maker who, having put the human in question by mimicking it with amachine, started to engage in other forms of taxonomic adventure.Orientalism was at its most vivid in the houses of Vichy and of Roulet & Decamps, where Japanese Ladies and Indian snake charmers were ranged alongside child acrobats and monkeys costumed as footmen.

The automaton's vaudeville counterpart proves that even the most convincingly mechanical act may be manipulating its audience with such expertise that the laughs and the claps seem to be automatic responses. Repeatedly, stage mesmerism or automatism burlesques the old imperial dream of power by demonstrating that agency is never very far removed from activity, and that what moves on command is likely to have some motivation up its sleeve. The creators of Swan Lake could not stop its prima ballerina from making her own interpellation--thewild outbreak of thirty-two fouettes in the climactic courtscene--thereby instituting for all future productions a truly vertiginous moment in which the ballet is beyond the control of anyoneother than the performer who has taken the center of the stage. Florence and Frederick (or The Theatre and the Factory)

In the early twentieth century, there is a reaction in the realms of technology away from the automaton as rarefied ornament, a movement to stop it merely performing and send it to work. This is where the robot makes its appearance. Some of the most insistent anxieties surrounding the automaton are transferred to it, while some new ones aregenerated from the changing cultural context. It is surely no coincidence that the rise of the robot fantasyfollows the demise of colonial slavery: the anthropomorphic machine promises an untroubled dream of power by offering the prospect of guilt-free slavery. Isaac Asimov, responding to some thirty years of anxieties developed since the turn of the century, was careful to equiphis fictional robots with "The Three Laws of Robotics" as a programming code guaranteeing locked-in limits to human/robot power play:

1. A robot may not injure a human being, through inaction, allow a human being to come to harm.

2. A robot must obey the orders given it by humanbeings except where such orders would conflict with the First Law.

3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

Nevertheless, the fantasy is unsustainable. Asimov's more mature stories set up moral dilemmas arising from inevitable tensions at the robot/human boundary. Karel C+apek's RUR (1918) dramatises the fantasy of the perfect slave worker only to overturn it by staging a robot revolution and industrial take-over.

So when, offstage, the robot goes into thefactory, it does not look like one of C+apek's human look-alikes. It is much safer that way, from the point of view of the cultural imaginary.Charles Babbage led the movement away from anthropomorphic andperforming automatons and towards purely functional,industriallyoriented automation. Not coincidentally, he was also known for pursuing something of a vendetta against street performers, whom he would chase out of earshot when they played near his house.

The growing commitment to Frederick Taylor's principles of scientific management after their publication in book form in 1911 meant the increasing schematisation of factory work, in accordance with principles that took the machine as a model for the worker, rather than vice versa. Taylor came to management via engineering. As an engineer, he developed something of an obsession with trying to achieve the optimum functioning of a machine by examining each of its parts to see whether any adjustment might benefit the overall sequence of operation. With a logic unencumbered by sociological or psychological forms of awareness, Taylor set out to improve the functioning of the workers withexactly the same methodology. The resulting principles of scientific management repeatedly emphasised the need to divorce doing fromplanning. The worker who had been "scientifically selected" according to his "personal coefficient" for the work to which he (Taylor's worker isalways "he" or "a man") is best suited, is likely to be quite ill-equipped to understand the science of the work process. For example: Now one of the first requirements for a man who is fit to handle pig iron as a regular occupation is that he shall be so stupid and so phlegmatic that he more nearly resembles in his mental make-up the ox than any other type. . . . Therefore the workman who is best suited to handling pig-iron is unable to understand the real science of doing this class of work. He is so stupid that the word "percentage" has no meaning to him.

In other words, the worker in a very real sense does not "know" what heis doing. If he is "a high priced man" (one who wants to increase hiswage), he is well advised to make himself as much like a robot as possible: Now, hold on, hold on. You know as well as I do that a high priced man has to do exactly as he's told from morning till night. . . . [Here the worker is introduced to his new manager.] When he tells you to pick up a pig and walk, you pick it up and you walk, and when he tells you to sit down and rest, you sit down and rest. You do that right, straight through the day. And what's more, no back talk. If Taylor could only have installed his workers with the program for Asimov's three laws, the industrial relations problem might havedissolved over night.

Such instructions point to a radical distinction between "performing" an action and merely executing it. "A creature of our species," writes Maine de Biran in the 1870s, "having the exterior forms of the human, yet who neither knows nor is in possession of himself, does not existfor himself."

For de Biran and many other late-nineteenth-century philosophers, the definition of the human hinged upon the property of free will, which in turned hinged upon the capacity to governmovement with volition, so "the sense of the self is neither arrestednor suspended as long as the exercise of the will power or the freeforce acting, sui juris, to determine the locomotion of the body. . . has not been suspended".

Something like robotisation was the reportedexperience of many workers who experienced scientific management as theredesign of their every movement in order to eliminate wasted effort, maximise speed, and optimise precision. John Dos Passos gives a literaryimpression: "reachunder, adjustwasher, screwdown bolt, shove in cofferpin,reachunder, adjustwasher, screwdownbolt, reachunderadjustscrewdownreachunderadjust, until every ounce of life was sucked off into production and at night the workmen went home greyshaking husks."

The theme of loss of agency by sorcery, which had fascinated nineteenth-century theatre in classic works such as Swan Lake and Coppelia, took on a new guise, with the sorcerer being replaced by the factory manager and the bewitched creature by the worker as automaton. According to the leader of the 1913 Renault workers strike against Taylorism, scientific management "eliminated, annihilated and banished personality, intelligence, even the very desires of the workers, from the workshops and factories."

Scientific management saw the strike as a spoke in the wheel. A 1929 election poster for the Conservative Party in Britain using precisely this image epitomises the abstract thinking of scientific management and its anxieties about the lack of fit between worker and machine. The residual determining power of the workers is not an energy but asymbolic obstruction to the machine, which is seen as the only true moving force.

In its Dionysian heritage, the theatre is the domain of wild life and carnival transformation, the antithesis of the proper destiny of the automaton that is, ultimately and in spite of its ornamental pretentions, the factory. When the juvenile lead in 42nd Streetis told by the director: "Ya supposed to be a dancer. All ya need is a coupla licence plates and ya look like a model T Ford," it's clear that the guy is in the wrong place.

Yet between the 1890s and the 1930s, the theatre and the factorydevelop a curious double act that veers between stand-off and seduction;"The theatre demonstrates a persistent fascination with the factory, by which it is both tempted and repelled. And in consequence, the role of the performer, especially in vaudeville and cabaret, veers unstably between the machinic and the carnivalesque. Vaudeville was an escape route--sometimes the sole escape route--for those who couldn't stand routine- bound work. To swap the factory for the theatre was to exchange functionalism for folly, the daily grind for the nightly crisis, a regular subsistence for an economy of feast and famine. Scientific management marks a polarity: this was its conscious aim, for it sought to achieve an absolute of schematisation. If Frederick Taylor has an antithesis as a manager, it is Flo Ziegfeld. One of Ziegfeld's earliest influences was Buffalo Bill's Wild West Show, described in the 1880s as offering a vision of "a wild untamed order of society."

Ziegfeld admitted to a series of childhood fantasies that began withhis riding out of town with the Wild West Show. In adult life, itis as though he saw the economic environment of his times rather as Buffalo Bill saw the landscapes of the wild West: somewhere in whichadaptation meant a constant process of outwitting the otherwildlife--human and animal. In this environment, nothing could be moreuseless than a formula or instruction--or, for that matter, a principle.It was a matter of assessing any situation for the hazards oropportunities it might yield, sizing up the risks, and usually deciding to take them. The 1907 Follies program announced the characters "in the disorder of their appearance."

Ziegfeld's disorderly, capricious, and wantonly extravagant approach to putting a show together was the despair of his collaborators. He raisedthe principle of waste to an art form, on one occasion making a snapdecision to scrap a number on which $27,000 had just been paid out forthe construction of a giant-gold mesh handbag. Theatres like Ziegfeld's were places in which it was possible to make your name and your fortune, to reclaim some of the human agency that the factory was vampirically sucking into its machinery. While the mass of workers were ground down by the tyrannies of numerical control, Ziegfeld could squander $20,000 on gratuitous costume changes for the chorus, and Mae West could get up on stage and propose that numbers, like everything else, just needed to be seduced: "I learned early that two and two are four and five will get you ten if you know how to work it."

Mae West and countless other vaudeville artists and entrepreneurs knewhow to work it. The Mass Ornament A third figure needs to be introduced into the dichotomous picture of Taylor versus Ziegfeld, and this is John Tiller. Tiller was a Manchester businessman. He owned a cotton mill and came from the heart of factory culture, being born in one of a row of production-line terraced houses in a prototypical working class suburb of Manchester. He was a skilled business manager, but his real interests lay in theatre management and he started giving dance classes. After working with various local amateur groups, he scored an exceptional success with an item that he produced on commission for a pantomime in Liverpool. This was in 1890, when Taylor was only just starting to get a reputation for his ideas about scientific management and had yet to publish them, but Tiller appears to have been discovering them for himself. Knowing nothing about the techniques of dance, he set his mind to technicalities: his "abiding irritation with sloppy dance routines" led to an exclusive concentration on precision and synchronisation. Dancers, like cotton, could surely be put through the mill (which is where his training sessions took place).

Something equivalent to Taylor's scientific selection, based on measurement, was involved in his choice of personnel. The four children who were to feature in his first showpiece were to be as near identical as possible in height and build. In order to emphasise the idea of multiple copies, he gave them all identical dolls and choreographed ageometrical pattern of synchronised movements, in which they were drilled from early morning until late at night, frequently being left soexhausted that their parents had to carry them home. The precision routine of "The Four Little Sunbeams" was, from a theatrical point of view, pure electricity. The audience wanted more, and Tiller set about producing it. Once the prototype had been successfully tested, he set about mass production, mounting sell-out shows advertised as presenting "200 trained children."

These children were from theworking-class families of Manchester, andTiller gave them a form of work that closely parallelled that of the factory, while offering a range of modest bonuses. Dancers were sought all the time, and when girls turned up for auditions (which involved showing their legs and teeth, but not demonstrating any movement), they were asked to bring friends and sisters. After the first ten years or so of his enterprise, Tiller complained that he had run out of suitable girls in Manchester and opened an office in London from which a new recruitment campaign was run. "Auditions" were held in this small office, with the desk remaining in the center and occupying most of the floor space. Dance skills were not being tested at this stage, but once recruits were signed up, they embarked on an intensive drilling program that equipped them for membership in what were becoming internationally marketable troupes. It was not long before the recruitment process in London was being duplicated from offices in New York and Paris, while the headquarters in Manchester were moved to larger and grander premises. Dozens of new troupes were sent out on the music hall circuits, and some of these--notably the Palace Girls, the Plaza Girls, and the La Scala Girls--gained strong reputations in their own right.

The so-called training was really very like what happened in thefactory: new recruits were placed next to an old hand and told to copy their every movement as closely as possible. As the numbers grew and Tiller found he could work several troupes concurrently, each group had a Head Girl--the equivalent of a foreman--whose job was to keepdiscipline and constantly monitor the quality of work. During thepantomime season, they would do two shows a day, eating a staple diet ofegg and chips. There was a production-line approach to stage preparation. The girls queued, a dozen at a time, to have their faces covered at high speed with pan-stick; there were twelve girls to a dressing room, sharing one bucket of water for cleaning up after work. A make-up for the legs known simply as "wet white" was found to be cheaper and more practical for dancing than stockings, but when they came off stage, the dancers had to queue to wash it off in the bucket. Pan-stick was removed with lard, which was then wiped off with a towel. Grime, sweated labor, and poor working conditions were things the Tillers shared with workers in the factory, but they earned about a third more. Three pounds. The formula was "a pound for me Mum, a pound for me digs, and a pound for me."

The pound-for-me was a vital margin of surplus and a margin of freedom.While the stage act itself turned them into a collective automaton, the theatre offered a way of life whose irregularity fostered individual ingenuity. Most significantly, it offered geographic and social mobility. As the Tiller brand name took off, it led to an export trade, and the troupes found themselves posted to various parts of the globe: they worked for Ziegfeld in New York, at the Follies Bergères in Paris, at the opera house in Monte Carlo, and in Berlin cabaret during the Weimar years.

In Weimar Germany, they were immediately read through the lens of industrial politics. Paul Simmel drew a cartoon of Tiller dancers coming off a Ford production line. The automaton association was also picked up in a parody by Grit and Ina van Elben, acting as "endgirls" to a line of articulated wooden cut-outs. Seigfreid Kracauer wrote one of his most influential essays on The Tiller Girls, mistakenly assuming they were "products of American distraction factories."

Coining the term "mass ornament" as a lable for these "indissoluble girl clusters," Kracauer delivered a concentrated dose of Weimargrimness: The hands in the factory correspond to the legs of the Tiller girls. . . . It is the legs of the Tiller girls that swing in perfect parallel, not the natural unity of their bodies, and it is also truethat the thousands of people in the stadium form one single star. Butthis star does not shine, and the legs of the Tiller girls are an abstract designation of their bodies.

To Kracauer, the Tillers expressed "the Ratio of the capitalist economic system" (81)--ratio being an abstracted form of reasoning that does not encompass the organically human. There is something fundamentally mistaken, though, in Kracauer's attempt to take a majorcultural reading from the Tiller girls. Their regimented presentation took on connotations in Weimar Germany that were much more literal than those which attached to them in Britain. This is not to suggest that Kracauer was wrong in identifying the machinic qualities in the Tiller choreography, but the association of the machine with the massing of social and political forces into vast abstract patterns is specific to the Weimar context. Kracauer's explicit concern is to define the qualities of an epoch by identifying the ways in which public taste has changed in "the domain of body culture" (75). Without knowing the already extensive genealogy of the Tiller phenomenon, he associates these changes in taste with its arrival on the scene.

The Tiller Girls reflected both changes and continuities in the domain of body culture from the 1890s through to the 1970s and one of their most remarkable qualities was their capacity to preserve certain forms of consistency in their presentation by maintaining a level of parodic detachment from epochal change. During both world wars they affected kitsch versions of military and naval uniform, but the routines they performed were not substantially different from those originally designed for turn-of-the-century chorus lines costumed in layers of frilly petticoat. One of the pleasures of preparing this paper was reading Doremy Vernon's Tiller's Girls, a history based on interviews with over 200 women who had been among them, including some from the original troupes. Vernon's account leaves an impression quite opposite to that created by Kracauer: that there was no real loss of agency for anyone who joined the kick lines. The routine of the stage work was counterbalanced by constant unpredictabilities in the situations they encountered through working in so many different theatres, with so many different shows, in so many different parts of the world. Over several generations, countless women joined the Tillers as an available alternative to a working life in factories or shops, and left some years later with a completely different range of prospects. This was not always the case, of course; many returned to their original way of life and faced cruelly restricted prospects, but the experience itself seems to have been, by and large, remembered fondly. When the 100th anniversary of the first Tiller Girls performance took place in April 1986, over 300 former members came to the reunion to exchange anecdotes and photographs, demonstrating what Julia Parker calls "a tight cameraderie" in the troupes.

If the Tillers were production-line workers, they were also both the machinery and the product, in a parodically impossible configuration: an Escher loop, perhaps (to reinvoke Hofstadter) around which energy travels through three conversion points. They proved that the life ofthe commodity--especially when it is part of an export trade--is a great deal more various than that of the worker. To be a living commodity might be regarded as equivalent to complete loss of agency, but to be a living commodity with all your wits about you, protected from the base-line exploitation of prostitution by the standard that ensured your market value, was a situation with variable possibilities.

Anxieties about transferred agency are developed through a dichotomous logic that counterposes the human and the machine, the agentic and the automatic, the vital and the functional, as mutually exclusive equivalents. As Hofstadter illustrates, the logic works exquisitely when it is taken into the realms of pure equation. Either Babbage or Turing must be the program; either Turing or Babbage must be the programmer. It is the very equiavalence, though, that suggests the possibility of exchange. The exchange is pure and total: an instantaneous switching of positions. The same purity of logic is demonstrated in Taylorism, where it poses as the logic of science and mathematics, although it is indistinguishable from the logic of sorcery expressed in such works as Swan Lake and Coppelia. Historically, there is a deep complicity between sorcery and mathematics, a complicity that arises from the process of dealing in abstract configurations (or, perhaps, configured abstractions). Kracauer observes that the Enlightenment was also the age of the fairy tale and that "eighteenth-century reason recognised the reason of the fairy tales as its equal," but he draws a line between such reason and "the Ratio of the capitalist economic system" that is emptied of vital and organic elements, and so emptied of "magic."

As an example of the mass ornament, the Tiller Girls are seen asrepresenting commitment to a form of abstraction that is essentially hollow, a vacuum. Kracauer sees the recovery of cultural vitality as something to be achieved not by running away from the mass ornament, but by moving through and beyond it. What his own logic misses is the power of the vacuum itself, so he leaves Hofstadter's equation incomplete, with the notion of transfer missing. By the laws of physics, a vacuum seeks to draw into itself whatever there is of substance or energy in its surroundings. The "mass ornament" may be the demarcation of an alchemical space, a site in which energy and agency will seek a resurgence.

Finally, though, the anxiety that comes to a head in cultural readings of the Tiller Girls is that which hinges on the equation of agency with individuality. As we make the transition from a machine model of technology to a networked one, new ways of thinking about agency open up. The electronic artist Stelarc is currently engaged in a series of internet linked performances that arrange for the body of the performer to be moved by someone situated remotely and selecting the movements from a body map on a computer screen. From such experiments, Stelarc spins a wild fabric of hypotheses:

Imagine a body that can perform an action without memory, a body that can make a motion without knowing that it will carry it out, an action without any expectation. . . . Consider a body driven by multiple agents, remotely situated and spatially separated. . . . A problem no longer of having a split personality, but rather a split physicality. .. . In other words the body becomes a host for another agent. Electronically coupled bodies could then extrude agency from one body toanother body in another place.

Such propositions would have caused the nineteenth-century expert on movement and motivation to blow a gasket. Throughout his work, Stelarc has been exploring the idea that agency and intelligence are not properties of persons but "in the system." Here is the glimmer of a newparadigm shift that even Hofstadter didn't foresee. If Babbage andTuring has hit upon the concept of extruded agency rather thantransferred agency to account for their experiment, there might have been no argument. Yet the model of transferred agency and its attendantanxieties show a remarkable propensity for retaining their formulationin the face of the paradigmatic earthquake foreshadowed by Stelarc.Arthur Kroker writes of the electronic era as an age of dark sorcery, in which the "possessed individual" is emptied of agency by the invading vampiric powers of a machinic collective known simply as technology, "not technology as an object we can hold outside ourselves, but technique as us, as a grisly sign of the possession of body and mind.