a Automatism

The word “automatism” appears pervasively in writing at the end of the nineteenth century, although the word was of fairly new coinage at the start of the Victorian period.⁹ Psychophysiology made “automatism” an essential concept in explaining behavior, and by the end of the period it had become everyday psychological language. The new psychology produced a transformed conception of humans, understanding them as part of a complex causal system. The belief that a great deal of human behavior is determined by automatic or unconscious processes was accepted and discussed widely in both specialist and popular forums.¹⁰ Reflex action became an established physiological fact by the first part of the nineteenth century, and Marshall Hall’s theory of spinal cord reflexes was extended to reflex actions of the brain.¹¹ Experimental research on reflexes confirmed a separation between sensation and consciousness. Thomas Laycock made the case that brain reflexes function “automatically and mechanically, without the intervention of any sensation, feeling, thought, volition, or act of conscious mind whatever.”¹² Automatic functions, sensorimotor responses, made what had been earlier understood as conscious mental processes into purely mechanical ones.

The prominent psychologist W. B. Carpenter coined the phrase “unconscious cerebration’” in the fourth edition of his Human Physiology (1852) to describe reflex and other modifications of thought that are outside of conscious awareness. In Principles of Mental Physiology (1874), Carpenter writes: “a large part of our Intellectual activity … is essentially automatic, and may be described in Physiological language as the reflex action of the Cerebrum.”¹³ Reasoning, complex cognition, and creativity can all occur as automatic reflex action. If the dominant split in thinking about the human had been physical versus mental, or body versus soul, a new organizing division entered the discussion: conscious/volitional versus automatic.

Some who acknowledged the importance of automatic brain functions nevertheless held tenaciously to an account of a conscious directing will. Although Carpenter accepts that automatisms – unconscious cerebration – result in complex behaviors, such as problem-solving or the composition of poetry and symphonies, he insists, here in an 1875 lecture entitled “Is Man an Automaton?,” that automatic functions are “directed and controlled by the will; – the Ego willing the result, and leaving it to the automaton to work it out.”¹⁴ Carpenter acknowledges the role of automatism, but places it in service of a determining self. Over time, he argues, what were once volitionally directed behaviors become automatic “habits”; but again for Carpenter the will guides those automatisms, at least in a healthy mind.

Many agreed with Carpenter that nervous reflexes play a predominant role even in creative activity. E. S. Dallas’s The Gay Science (1866) proposed that all creative work is produced unconsciously. Oliver Wendell Holmes asserted: “The creative act is not voluntary at all, but automatic.”¹⁵ Robert Louis Stevenson famously attributed the story of Dr Jekyll and Mr Hyde (1886) to his metaphorical “brownies” who work outside of his awareness. But others opposed the idea that creative works, regarded as the highest human achievements, could be produced without a directing consciousness. Linda Austin’s chapter in this volume situates the reception of Anthony Trollope’s work in a prominent line of criticism resistant to the role of automatism in creativity.¹⁶ A version of that view is evident in Max Nordau’s contention that genius overpowers automatisms, thereby preserving the spontaneity of conscious thought. “Even the highest centers are not purely human as long as their perfection does not reach beyond automatism,” he writes.¹⁷ Austin shows how criticism from Trollope’s time into the twentieth century adhered to a model of conscious intention in creativity, and demoted Trollope’s writing because of his mechanized writing habits.

The concept of automatism was used by the new psychology in diverse ways. As experimental psychology was established, research on automatic processes of mind moved into physiology laboratories, where precision instruments registered responses and involuntary actions. In early psychiatry, automatisms were connected to insanity and understood as pathological or animal-like behavior, a resurgence of lower and earlier developed functions of the brain. Evolutionary models of mind reserved volition, rationality, and consciousness for what was deemed the further evolved brain, contending that automatisms were more dominant in nonwhite races, women, and the lower classes than in white European men.

The theory of human automatism that best captured the Victorian imagination was Thomas Huxley’s conscious automatism. Huxley argued, to great notoriety at the 1874 meeting of the British Association of Science, that humans are automata. According to Huxley, an inefficacious consciousness takes itself to be in control, when in reality it merely accompanies the automatic, unconscious production of behavior. That is, all human behavior is automatic. In explaining his theory, Huxley draws explicitly on the history of the production of mechanical automata. He invokes the admired eighteenth-century automata maker Vaucanson – who remained an often-mentioned touchstone of achievement in the Victorian period for his complex mechanisms – discussing a soldier with neurological damage, who apparently performs complex actions with no conscious awareness of his doings. Is the soldier, Huxley asks, “in the condition of one of Vaucauson’s [sic] automata – a senseless mechanism worked by molecular changes in his nervous system”?¹⁸

Huxley’s polemical presentation of human automatism was debated extensively for the remainder of the century. My chapter in this volume examines the uptake of the conscious automaton debate in the broader culture, considering the implications of the view for thinking about consciousness and literature. In Chapter 6, Tyson Stolte demonstrates the influence of theories of human automatism on James Thomson’s thought and his portrayal of agentless characters in a deterministic world in the poem “The City of Dreadful Night” (1874). Stolte argues that Thomson’s departure is that he does not allow his characters even the illusion of the efficaciousness of consciousness. Yet, Stolte shows, Thomson’s readers were able to interpret that poem as expressing a view closer to Carpenter’s directing will than to Huxley’s conscious automatism. George Levine, in his contribution, shows that Darwin’s theories have been taken by many, including his contemporary Samuel Butler, to entail the Huxleyan view that people are conscious automata and that intelligence and intention play no roles in a mechanistic natural world. Levine demonstrates that instead aspects of Darwin’s writing undercut a thoroughgoing automatism and a world emptied of mind and volition.

b Volition, the Law, and Hypnotism

A persistent note in Victorian writing is that automatist theories radically diminish the role of volition in human action, and in so doing give rise to anxieties about moral, religious, and legal responsibility.¹⁹ If behavior is determined, nonvolitional, and automatic, then it seems that individuals cannot rightly be held responsible for their actions. Discourses of automatism shaped ideas of criminal responsibility and influenced legal proceedings and rulings, as Roger Smith and Joel Peter Eigen have demonstrated in important monographs.²⁰ Courtroom defenses and verdicts had recourse to the new defense that criminal acts had been performed by perpetrators in unconscious, automatist mental states, such as while sleepwalking. Roger Smith’s contribution to this volume shows that noted criminal trials involving automatism from the 1850s on worked as public theater and helped crystallize the terms of the more abstract philosophical debates on automatism in the 1870s. Smith argues that the ostensibly descriptive language of automatisms imported moral and emotional, that is, evaluative meanings, even while questions regarding automatism and volition worked to shift attention away from the social conditions that, in fact, led to crime.

Even at the start of the Victorian period, mesmerism and, in 1843, James Braid’s more scientifically accepted hypnotism (physiologically produced trance states), had established the workings of automatisms in human action. A person (often understood as weak-willed) succumbed to the mesmerizer’s control or performed actions while in an alternative mental state. Discussions of mesmerism also raised concerns about the role of volition in action. Could a mesmerist control the actions of a person, rendering them a machine without awareness or moral sense, a human automaton? The idea of the mesmerized or hypnotized human subject continued to figure in public discussion and literary productions throughout the century, with concerns raised regarding the possibility of automatized subjects driven by the will of the hypnotizer. Popular literature obsessively produced stories of criminal acts performed in a hypnotic state, and of woman rendered automata, perhaps most famously in George du Maurier’s Trilby (1894). Arthur Conan Doyle exploits that scenario in his 1887 novella The Parasite, where his scientist narrator explains that his fiancé had “been reduced to a condition of automatism…. A person at a distance had worked her as an engineer on the shore might guide a Brennan torpedo.”²¹ Doyle imagines that a human could be guided like a remote-controlled weapon, her “nervous mechanism” externally controlled by the will of another. Worries about hypnotic suggestion in crime were not restricted to fiction, but played a role in both medicine and criminal cases at the end of the century.²² A somewhat subtler, but conceptually related idea appears in George Gissing’s late-century fiction, as in The Nether World (1889), where characters are said to be “driven automaton-like by forces they neither understood nor could resist.”²³ For Gissing, the force is no longer externalized in someone’s will, but is instead environmental – social – and perhaps hereditary; the environment rather than another agent compels the crime. Shuhita Bhattacharjee’s chapter in this volume examines the depiction of the mesmerized person in Richard Marsh’s little-studied popular fiction beyond The Beetle (1897), arguing that Marsh employs the figure of a female or impoverished subject rendered an agentless automaton to make manifest social and legal injustices and power imbalances.

c Automatism in Occultism and Spiritualism

Theories of automatism were rejected by some not only because they threatened beliefs about human agency, but also because they diminished human exceptionalism. Automatism made the mental a part of the natural world and put humans on a continuum with animals and, for a few commentators, even machines. Those who resisted what they regarded as a reductive materialist view sought to reassert the reality of the soul and mental capacities that went beyond what naturalist science would accept. Outside of orthodox religion, spiritualist and occultist movements presented evidence of communicating spirits and an unseen realm that could in some respects be known by the human mind. Psychologists such as Carpenter and Morton Prince, on the other hand, drew on theories of automatism to debunk claims of communication with the dead and other psychical phenomena. They explained the phenomena by appeals to unconscious cerebration, the power of suggestion, hallucination, or mental pathologies.

But theories of automatism were increasingly adopted by those who sought to establish the reality of extraordinary mental powers or to provide evidence of the continuation of a nonphysical self after death.²⁴ Spiritualist mediums were said to enter into automatist states in which they were able to contact the dead or access alternative planes of reality. The medium’s body (often a woman’s) in an automatized trance could deliver spirit communications.²⁵ Automatic writing, widely discussed late in the century in medical and psychological journals, was declared by some to be evidence of contact with another world. For Frederic Myers at the end of the century, “automatisms” became a central category of explanation of extraordinary mental phenomenon. Automatisms for Myers permit access to subliminal or secondary consciousnesses. He describes these as “nunciative,” that is, message-bearing, automatisms.²⁶ Nunciative automatisms may deliver “supranormal” knowledge coming from sources outside the person, but more often they are “messages communicated from one stratum to another stratum of the same personality” (524). Myers’s automatisms return control to “the ‘self’ which supplies the initiating power” (523), a self that resides outside of normal consciousness yet is not pathological. Aren Roukema’s chapter in this volume explores themes of psychic power and mind control in the context of scientific racism. Roukema examines popular occultist fiction that depicts psychic automatisms in ways that challenge Victorian evolutionary science’s racial hierarchies that position white European men at the top of the evolutionary ladder. The works Roukema analyzes instead situate psychical superiority and advanced mental power in nonwhite characters.

d Automatism, Race, and Scientific Racism

As Roukema’s chapter shows, proponents of race science exploited evolutionary theory to argue that less evolved – that is, nonwhite – brains were dominated by unconscious, instinctual, and automatic functions. That nonwhite people are automaton-like and lack agency became a pernicious leitmotif in the writing. Long Bui has shown in his Asians as Automata that Chinese laborers in the nineteenth century were denied personhood and demoted to labor machines: “Asians act and behave like numbed automata bereft of deep feeling, spontaneous thought, and human consciousness.”²⁷ Those ideas recur frequently in British travel writing in the period. The anonymous writer of a piece on Hong Kong, for example, writes that “reason and physiological knowledge fail in explaining the constitution of the Chinaman, and you can scarcely help wondering whether he is part of the organic creation at all, but not rather some marvelous automaton cunningly devised by Providence for the special benefit of the European.”²⁸ The racist Eurocentrism is so extreme as to read like satire, in its claims for white ownership of reason, scientific knowledge, God’s partiality, and even biological matter.

Similar strands of scientific racism are embedded in much of the psychological research in the period. Meade Bache, in a paper on “Reaction Time with Reference to Race” (1895) in The Psychological Review, argues with convoluted logic that it can be demonstrated that “the negro is, in the truest sense, a race inferior” because of his superiority in “the quickness of his automatic movements.” He continues: “The negro is, in brief, more of an automaton than the white man is.”²⁹ As Chris Dingwall’s chapter in this volume shows, such self-interested constructions permitted white people to understand themselves as being in sole possession of rationality, as autonomous and self-determining, while representing Black people in opposing terms as automatic and externally controlled. Examining developments in the United States following the abolition of slavery, Dingwall traces the character of Topsy from Harriet Beecher Stowe’s Uncle Tom’s Cabin’s representation of the slave girl as mechanized human to the subsequent factory production of Topsy dolls, arguing that the cultural uses of Topsy point to the “racist iconography” of automata in the nineteenth century.

a Race and Mechanical Automata

That same racist iconography is on display in Victorian mechanical automata, both in exhibitions and as commodities. A succession of orientalist automata, with “exotic” names such as Hajeb and Ajeeb, reproduced Wolfgang von Kempelen’s famous eighteenth-century automaton chess player, the Turk. Ajeeb, described as a mystical “Moorish figure,” was another sham mechanical chess player (like Maelzel’s Chess Player analyzed by Poe). Ajeeb was exhibited at the Royal Polytechnic Institute in 1868 and displayed at London’s Crystal Palace into the 1870s, before being toured across Europe in the following decade.³⁰ Contemporary articles raised the possibility of occult or mesmeric explanations for the automaton’s skills, reinforcing cultural narratives regarding the irrationality of the “East.”³¹

Mechanical Black and Asian automata were widely available as collectible items in the later decades of the century. In his discussion of the gender of automata, M. Norton Wise has briefly linked the increase in production of Black automata to evolutionary theories that asserted the greater automatism of the nonwhite mind, as well as the female mind.³² Louis Chude-Sokei, in The Sound of Culture, examines the connection of Black bodies to laboring machines in the United States: “In the antebellum period the relationship[s] between blacks and machines were so established that there were a great many ‘automaton “Negroes”’ produced in America.”³³ In Chapter 2, Edward Jones-Imhotep and Alexander Offord observe that the many Black automata created in the nineteenth century carry out distinctly nontechnological activities: fruit-selling, banjo-playing, smoking, and singing. Jones-Imhotep and Offord question why historical research has not paid greater attention to the pervasiveness of Black automata, and contend that turning to discourses of scientific racism to explain the pervasive production of Black automata is to replicate the white gaze. Instead, they argue, Black automata should be understood as racist objects, constructed to create and sustain a myth that Black people and technology are not compatible.

b Thought Problems, Spectacles, Exhibitions, and Commodities

Jean-Claude Beaune, in an influential article on the history of automata, argues that mechanical “automata represent the dream, the ideal form, the utopia of the machine.”³⁴ Automata are alluring because they appear to be alive and to act on their own. But, as the racist automata of the nineteenth century make clear, mechanical automata have many meanings. The particular forms automata take, the dress they wear, the painted color of their skin, the activities they engage in, where and how they are sold and to whom, all matter to their meanings. In the Victorian period, mechanical automata continued to raise abstract philosophical problems and to entertain, but they also increasingly embodied desires and reflected particular social meanings as they were enmeshed in an accelerating commodity culture.

The popular stage magician Jean-Eugène Robert-Houdin describes in his Memoirs (1859) the great delight he took in constructing entertainment automata – his “progeny.”³⁵ Stretches of his autobiography dwell, with a curious affective intensity, on the unparalleled happiness that he experiences when inventing a writing/drawing automaton; he discovers that he has “unconsciously produced an exact likeness of” (202) himself – the mental automaton creating the mechanical one. Setting his automata in motion for the first time, the initial question he puts to it is a Frankenstein-like: “Who is the author of your being?” (202). Robert-Houdin associates his invention of automata with reproduction of self, but also with a deity-modelled control. There is another curiosity in his account. He explains that he strived to make his automaton’s mechanism silent, attempting to “imitate nature, whose complicated mechanisms work almost imperceptibly” (205). But because of that lack of noise, viewers undervalued his writing/drawing automaton, taking the silence to indicate simplicity of construction. It was only when he reworked the mechanism so that it produced a “whizzing sound” that audiences came to admire the complexity of his creation and his talents. While it has been common to understand mechanical automata as vehicles allowing thought experiments on human mechanism,³⁶ Robert-Houdin’s story suggests that mechanical automata appeal not only because they draw attention to similarities between humans and machines, but also because they underscore differences, refusing to represent the reduction of the biological to the mechanical.

Automata played a role in exhibition and entertainment culture in the Victorian period, as they had in the late eighteenth century. Fraudulent chess-playing automata continued to draw crowds in the nineteenth century. They were on display at fairgrounds, at technological exhibitions, and in galleries of practical science. Newspaper articles reported on the crowds gathered around music-playing automata at expositions such as London’s International Exhibition of 1862.³⁷ Automata were used in stage spectacles, in magic displays, or in shows in which performers mimicked mechanical automata. Stage automata befuddled audiences, raising questions about the differences between humans and machines and about the ultimate capabilities of mechanism. John Tresch argues that automata on display left viewers uncertain as to whether uncanny forces were involved or whether the actions were solely mechanical: “the performances made audiences hesitate, oscillating between mutually contradictory interpretations of the scenes before them.”³⁸

Research specifically focused on mechanical automata in the second half of the nineteenth century has primarily looked at the French context. Christian Bailly’s Automata: The Golden Age collects images of elaborate mechanical automata produced in Paris in the last decades of the nineteenth century – dancers, musicians, conjurers, acrobats, white children involved in leisure activities, snake charmers, and Chinese opium smokers. Monkeys dressed in human attire, doing human activities, abound, suggesting that these automata embodied questions not only about the boundaries between humans and machines, but also humans and animals. Automata were increasingly a collectible commodity. Automata makers, working in their own quarter in the Marais district, created aesthetically impressive, artisan luxury automata for collectors. (Bailly reports that some automata sold for the amount on which a less affluent family could live for a year.³⁹) Those costly automata were sold by jewelers, who worked with watchmakers such as the renowned Patek Philippe to produce them for well-to-do buyers. As cultural capital, those automata, exported around the world, hovered somewhere between art and collectibles.⁴⁰ They were not as unique, expensive, or difficult to produce (given machine-aided production) as the automata of the eighteenth century had been. But, as Kara Reilly has shown, luxury automata were distinctly different from the relatively inexpensive, mass-produced automata that became popular as toys in the later nineteenth century and common objects in homes.⁴¹

Reilly’s essay in this volume looks at automata that reproduce celebrity stage performers in the late nineteenth century. She shows that automata were important not only on the stage, but also as commodity collector items that allowed the public to possess a “living” version of the performers. Her history of collectible stage luminary automata gives insight into the construction of the idea of celebrity, strategies of marketing, and the acceleration of commodity culture at the end of the Victorian period. Technologies of mechanical automata developed over the course of the period, from clockwork, to steam, to electricity. Automata also began to incorporate new technologies of stored memory. By the century’s end electrical automata were widely used in advertising in shop windows. Automata presenting as imitations of humans helped make consumers into imitations, modelling identities and desires.

If the automata of the eighteenth century drew on mechanism as explanatory model for the biological and sought to reproduce (simulate) the physiological functions and appearances of the body, externally and internally (as in Vaucanson’s much-discussed duck ostensibly imitating the process of digestion), Jessica Riskin argues that in the nineteenth century “the ambition was not to simulate” physiology. Automata in the period “instead were artistic renditions of animal and human activities.”⁴² Attempts at mechanical simulation gave way to new technologies focused on abstract functioning rather than material replication. There were some exceptions. Étienne-Jules Marey, for instance, constructed automata, including a mechanical flying insect, to support the graphic representations of flight inscribed by his recording apparatuses.⁴³ Automata continued to play somewhat of a role in scientific exhibitions in the Victorian period. An 1841 article on “Automaton Figures,” reports on “Professor Wheatstone, who exhibited a speaking automaton at the Dublin meeting of the British Association for the Advancement of Science.”⁴⁴ Physicist Charles Wheatstone’s construction of automata influenced his later work on acoustical theory and the propagation of sound, important in the development of telegraphy.⁴⁵ His speaking automata (which copied not a human figure but instead the human vocal tract) later inspired Alexander Graham Bell in his development of the telephone. Charles Dickens’s “Full Second Report of the Mudfog Association for the Advancement of Everything” (1838), notoriously satirizing the published reports of the British Association for the Advancement of Science, pokes fun at the use of automata in scientific exhibitions in the section “Display of Models and Mechanical Science.”⁴⁶ Dickens’s men of science advance plans for the construction of an automaton park (well before the development of actual amusement parks at the end of the century). The park, for dissolute young nobleman, would offer mechanical old women and cab drivers to knock around, along with automata policemen to take the offenders into custody and automata magistrates to exonerate them the following morning. Dickens’s parody evinces his perceptive awareness that mechanical automata reflect social ideas.

Contemporary considerations of the history of mechanical automata connect the development of automata directly to industrial capitalism and the work of empire, situating the earlier emphasis on automata as aesthetic objects or entertainment as preludes to practical, profit-making industrial uses. Hermann von Helmholtz, in his frequently cited discussion of Droz and Vaucanson, admires the eighteenth-century creations because they “enrich the mechanical experience which a later age knew how to take advantage of.”⁴⁷ Industrial machines in his day are produced to “perform one service, but shall occupy, in doing it, the place of a thousand men” (190). David Brewster, in his discussion in Letters on Natural Magic (1832), sees the wonder-inspiring automata of the eighteenth century as paving the way for “purposes of higher import,” the factory machines of the nineteenth century: “Those mechanical wonders, which in one century enriched only the conjuror who used them, contributed in another to augment the wealth of the nation; and those automatic toys, which once amused the vulgar, are now employed in extending the power and promoting the civilization of our species.”⁴⁸ For Brewster, the mechanization of reason in Babbage’s calculating machine sits at the apex of progress, outdoing any previous mechanical virtuosity.

c Industrial Automata and the Beginnings of Artificial Intelligence

Invocations of industrial automata like Brewster’s and Helmholtz’s were common in nineteenth-century discussions of the factory and labor. “[T]he automaton became a symbol for human machine relations in the industrial age,” writes Adelheid Voskuhl.⁴⁹ The automaton is used to signal degraded, alienated labor, a dehumanized worker demoted to will-less mechanism, but also signifies a complex, mechanical system that, though not conscious or living, is understood as intelligent. In an instance of the former use, Marx refers to the factory worker as a “‘living’ automaton.”⁵⁰ Journalist William Cooke Taylor in his 1844 work on the factory system similarly claims that repetitive “mechanical processes” “degrade the workman into an automaton.”⁵¹ Taking the analogy in the other direction, Andrew Ure’s well-known description of the factory, quoted by Marx in Capital (1867), portrays the factory as “a vast automaton, composed of various mechanical and intellectual organs,” a description that gestures at the idea that a complex system of machinery can be regarded as possessing intelligence.⁵² Marx’s famous description of the factory system depicts it as an “automaton” that reduces the worker to a part of the system: “the automaton itself is the subject, and the workmen are merely conscious organs, co-ordinate with the unconscious organs of the automaton, and together with them, subordinated to the central moving-power.”⁵³ Making the automaton “the subject” seems to situate agency in the factory mechanism, “endowed … with intelligence and will” (401). But the description also, like Huxley’s, puts unconscious, nonvolitional automatism in control, consciousness existing in the workers, but entirely inefficacious. Simon Schaffer has argued that “to make machines look intelligent it was necessary that the sources of their power, the labor force which surrounded and ran them, be rendered invisible.”⁵⁴

Writing on factory industrialization, then, worked to further destabilize the distinction between mechanical automaton and living person, biologizing the machine and mechanizing the human, while diminishing the significance of consciousness.⁵⁵ Contemplation of the overlap and boundaries between the human and the mechanical took new shape in the second half of the period in response to innovations in industrial technologies. As Anson Rabinbach has shown, a transformed model of the human in relation to industrial production emerged. The laboring body was reconfigured as a thermodynamic system: “the working body was but an exemplar of that universal process by which energy was converted into mechanical work, a variant of the great engines and dynamos of the industrial age.”⁵⁶ That new conceptualization encouraged comparisons between the mechanisms of human and industrial automata and gave rise to dreams of increased human efficiency and economic productivity. Those constructions also fueled fantasies of competition between industrial and human automata.

Conceptions of the factory as an intelligent system, like discussions of the mechanistic nature of mental activity, directed attention to the question: What is intelligence? That question was magnified by the first efforts to create “thinking machines,” most famously in Babbage’s attempts to automatize intelligence in his plans for his Difference and Analytical Engines.⁵⁷ As the physical mechanism was accepted as increasingly important to mental activity, cognition was regarded as decreasingly mysterious and consequently something that could conceivably be achieved by a machine. As has been frequently observed, early endeavors at calculating devices together with deliberations on ostensibly intelligent chess-playing automata (though in fact hoaxes) were precursors to current AI and robotics.

The development of machines replicating processes such as perception, memory, and reasoning further suggested that cognition does not require a conscious living subject. The economist Alfred Marshall delivered to Cambridge’s Grote Society in 1868 a remarkable early paper, “Ye Machine,” in which he presents a detailed plan for unconscious machine intelligence. He envisions machines with “bodies” and “brains” that sense and respond, and even reason mechanically: “A Machine can be caused to receive from the outer world, in the form of pressure, light, sound, etc., impressions or modes of motion which shall effect changes in the … relations of its internal parts.”⁵⁸ Those changes are “sensations,” in his terms, which result in “actions” by a process mediated by simple “ideas.” Marshall drew on current psychophysiological models in describing his intelligent machine, which would carry out processes of association, develop automatic behaviors through repetition, and possess memory. The devices would over time come to have “different moral characters” (130) and be capable even of creative production.

At the end of the Victorian period, Nikola Tesla forecast the invention of autonomous, thinking automata:

Tesla is confident that this can be achieved, because he models the machine on himself: “I am an automaton endowed with power of movement” (184). Cognition, he argues, can be realized in multiple media; substance is not important: “Whether the automaton be of flesh and bone, or of wood and steel, it mattered little” (185). Such speculations, if largely hypothetical and at the time still closer to scientific romance than reality, worked to refashion ideas of the automaton, bringing human and automaton ever closer.

If machines can think and perceive, what is left that makes humans special? Similar to current worries about AI and chatbots turning out writing and art, the specter of machines capable of thinking and even of producing creative works occasioned both anxieties and satire. Richard Altick discusses the performance of the “Eureka machine,” a “poetry making machine” for composing Latin verses, which was on display at the Egyptian Hall in 1845. Altick notes that an advertisement in The Athenaeum announced that the instrument was “the nearest approach to a thinking machine ever yet produced.”⁶⁰ The supposition that a machine could produce creative works was strongly resisted by some, as were the arguments in psychophysiology that creative work originated in automatic brain processes. As early as 1844, in a piece on “J. Babbage’s” “New Patent Mechanical Novel Writer,” Punch Magazine satirized the possibility of machine-composed literature, while simultaneously critiquing what some characterized as the mechanical nature of novel-writing of the day. Ada Lovelace, in a much-quoted remark, claimed that Babbage’s “Analytical Engine has no pretensions whatever to originate anything. It can do whatever we know how to order it to perform.”⁶¹ On her account, the automaton cannot invent anything; she reserves that special power for human mentality. But the concern that automated intelligence would be able to reproduce the products of human creativity circulated in the culture. At the end of the century George Gissing’s overworked character Marian in New Grub Street (1891), seeing an advertisement for a “Literary Machine,” hopes for “some automaton to supply the place of such poor creatures as herself, to turn out books and articles.”⁶² Gissing condemns the relentless labor and the pace of writing for the literary market, but Marian’s fantasy is grounded in the assumed overlap of biological brain and nonliving mechanism.

Sally Shuttleworth’s chapter in this volume considers George Eliot’s fiction, particularly “Shadows of the Coming Race” (1879), in the context of the development of machines with cognitive capabilities, such as logic machines and machines that could detect, interpret, and record sensory input on the model of human perception. Shuttleworth argues that George Eliot is concerned that her literary realism cannot compete with machines that outstrip human abilities of sensing and observation in accurate representation of the world. In Chapter 3, Simone Natale contends that theoretical questions that later emerged in the study of human–computer interaction (HCI) allow a reconsideration of Victorian discussions of human interaction with automata that display seemingly intelligent behavior. Bringing nineteenth-century automata into dialogue with current AI, Natale demonstrates shared ground – if in different terms – between HCI and Victorian writing on automata.