In the 1950s, John Cunningham Lilly, a prominent neurophysiologist and section chief at the U.S. National Institute of Mental Health (NIMH), stunned the scientific community when he declared it would soon be possible for humans to establish two-way communication with dolphins. “Within the next decade or two,” Lilly proclaimed in his widely-read book Man and Dolphin, “the human species will establish communication with another species: nonhuman, alien, possible extraterrestrial, more probably marine; but definitely highly intelligent, perhaps even intellectual.” ((John C. Lilly, Man and dolphin (Garden City, NY: Doubleday, 1961), 11.)) Widely reported in the news media, Lilly’s ambitious project at his purpose-built Communications Research Institute was perhaps the most visible element of a much broader effort initiated by the Office of Naval Research to integrate dolphins into certain naval operations, which required the production of knowledge and techniques for communicating between humans and dolphins. ((See Forrest G. Wood, Marine mammals and man: The Navy’s porpoises and sea lions (Washington: Robert B. Luce, Inc., 1973).))
While Lilly drew from a wide array of conceptual tools to build his case for the scientific study of interspecies communication, I highlight the importance of Lilly’s radically expansive concept of the communication network, which was inspired by emerging computational techniques and cybernetic models of mind and brain. As new understandings of the network as a feedback-controlled system gained traction in the scientific community and in the major funding agencies of the Cold War, Lilly was able to redefine the problem of interspecies communication as an engineering and coding problem – a matter of developing circuits, interfaces, and programs to connect neural “biocomputers” across species lines.
Cultural histories of Lilly’s networks, and of the broader discourses and practices of military-funded computing and communication projects in the Cold War period which informed Lilly’s work, provide a compelling narrative that helps to explain the emergence and significance of Lilly’s human-animal-machine networks. ((See D. Graham Burnett, “A mind in the water: The dolphin as our beast of burden,” Orion, May/June (2010): 38-51. D. Graham Burnett, The sounding of the whale: Science and cetaceans in the twentieth century (Chicago: University of Chicago Press, 2012). Mette Bryld and Nina Lykke, Cosmodolphins: Feminist cultural studies of technology, animals and the sacred (New York: Zed Books, 2000).)) Through these narratives, Lilly’s own account of his network-building project can be critically analyzed and connected to the socio-past of the military-industrial complex and the broader structure of domination through which human societies exploit nonhumans. Yet, the danger here is that the imposition of this compelling but “pre-given” historical narrative – the socio-past of the Cold War institutional matrix in the U.S. as the prime mover that creates and determines the meaning of Lilly’s networks – may lead researchers to overlook the struggle between competing understandings of network media during this period. How might we retrace the development of Lilly’s interspecies networks in way that, as Gabrielle Durepos and Albert Mills put it, creates “new forms of telling that surprise and liberate us from the type of history that is an effect of strict historical conventions designed to impose the idea of an (ultimately) singular history”? ((Gabrielle Durepos and Albert Mills, “Actor Network Theory, ANTi-History, and Critical Organizational Historiography,” Organization 19, no. 6 (2012): 703-721._))The main intervention I make here is to reorient the history of Lilly’s networks around the unexpected. In particular, I highlight a series of shifts in the institutional, conceptual and technological basis of Lilly’s network-building practices at the end of the 1950s, when Lilly abandoned what I call his “neuroelectric networks” at the National Institute of Mental Health and began to build bioacoustic and psychopharmacological networks at his own Communication Research Institute. I emphasize the disjuncture between these two types of networks by highlighting the collision between conflicting meanings of network media at the CRI. Whereas the neuroelectric networks were shaped by predominant understandings of networks as architectures of centralized command and control, the networks constructed at the CRI in the 1960s articulated a conceptualization of the world that might be described as, following Andrew Pickering, “nonmodern” in the sense that the networks embodied a disposition toward the world as a set of entities whose dynamic interaction across the human/nonhuman divide troubles conventional modes of scientific knowledge production. ((Andrew Pickering, The cybernetic brain: Sketches of another future (Chicago: University of Chicago Press, 2010): 28.))
The deviations of Lilly’s networks from dominant understandings of network media were unexpected, but they cannot be reduced to chance or to Lilly’s principled rebellion against the military-intelligence establishment (though the latter seems to have been Lilly’s preferred reading). Lilly’s move from neuroelectric to bioacoustic networks at the end of the 1950s and the more general shift in the dispositions toward the world articulated by these networks, can be understood in terms of Lisa Gitelman’s concept of the “material meanings” of media, that is, the “nexus of cultural practices, economic structures, and perceptual and semiotic habits that make tangible things meaningful.” ((Lisa Gitelman, “Media, materiality, and the measure of the digital; or, The case of sheet music and the problem of piano rolls,” in Memory bytes: History, technology, and digital culture, ed. Lauren Rabinovitz and Abraham Geil (Durham, NC: Duke University Press, 2004): 203.)) In the 1950s, Lilly stabilized the material meanings of his neuroelectric networks by aligning these networks with cybernetic concepts, practices and technologies derived from military and intelligence institutions. These institutions were investing huge sums of capital into the development of massive “Big L” ((Paul N. Edwards, The closed world: Computers and the politics of discourse in Cold War America (Cambridge, MA: MIT Press, 1996), p. 107. As Edwards explains, “Big L” systems were computer-based communication and control systems developed for in response to the displacement of airplanes by nuclear missiles in the postwar era, which reduced warning time and demanded rapid means of mobilizing and coordinating offensive and defensive forces. As Edwards points out, SAGE, which was also known as Air Force project 416L, was the model for twenty-five systems (with project codes ending in “L”) built during the 1950s and 1960s to counter the threat of intercontinental ballistic missiles. For Edwards, Big L systems embodied closed-world discourse in “computerized command and control.”)) surveillance and communication networks to provide early warning of nuclear attack as well as new techniques of interrogation and brainwashing through the notorious “impossible experiments” ((Rebecca Lemov, World as laboratory: Experiments with mice, mazes, and men (New York: Hill and Wang, 2005). Alfred W. McCoy, A question of torture: CIA interrogation, from the Cold War to the War on Terror (New York: Metropolitan Books, 2006).)) with drugs, electrical brain stimulation, and sensory deprivation. The material meanings of Lilly’s networks at the NIMH, like those of the Cold War military-intelligence establishment, were organized through what Paul Edwards calls “closed-world discourse,” which refers to a durable assemblage of “techniques, technologies, metaphors and experiences” in Cold War engineering and politics which constructed the world as a “closed system accessible to technological control” and mobilized the computer as the embodiment of this vision. ((Edwards, The Closed World, 113.)) I suggest that Lilly’s experimental techniques for mapping, reconfiguring and programming the brain-as-computer through neuroelectric networks extended closed-world discourse, including its preoccupation with human-animal-machine integration and its metaphors of “enclosure” and “containment” for strategies of power – from geopolitics and weapons engineering to cognitive and cultural processes such as learning and persuasion.
While no other form of network media exceeded the scale of closed-world networks in this period, the material meanings of networks were still open to contestation, particularly on the fringes of the closed-world networks. As Gitelman demonstrates, when new alignments between cultural, perceptual, semiotic and economic elements of media gain traction within particular institutions or cultural communities, the material meanings of media and of their material substrates (e.g., electrical transmissions, psychotropic substances, sonic emissions) become destabilized and open to reconfiguration in new types of network media. Whereas closed-world discourse aimed to construct and maintain defenses around and between enclosures, Lilly’s human-dolphin networks constituted what might be called, following Edwards, “green-world” networks which strive for liberation from, or destruction of, closed-worlds and operate as platforms for (often spiritual) journeys towards “transcendent powers,” community, and “cosmic order.” ((Edwards, The Closed World, 310.)) I argue that Lilly’s work exemplifies the postwar attempt to build green-world networks from the materials, techniques and economic structures of closed-world networks. The destabilization of the material meanings of Cold War networks, and the unexpected distributions of communication and knowledge production which resulted from this destabilization, brings into focus the multiplicity of networks (neuroelectric, bioacoustic, psychotropic) and conflicting ways in which those networks attempt to integrate humans, animals and machines.
Neuroelectric Networks as Closed-World Networks
The dominant perspective of networks in the postwar period was articulated through what Edwards (1996) calls the “closed world discourse” of Cold War military computing, which helped sustain the belief that nuclear war would be “winnable” by building networks that capture “the whole world within its closed system,” anticipating “every possible contingency.” ((Edwards, The Closed World, 292.)) As Edwards argues, this vision of the world as a closed system in which entities and their actions would be contained in a calculative and predictive environment of continuous surveillance, intervention and feedback shaped the mid-century development and use of computers and communication networks. While Lilly had no direct connection to the closed-world of nuclear strategy, the neuroelectric networks he constructed at the NIMH in the 1950s emerged from what Edwards calls “cybernetic psychology” which conceptualized the mind and brain as feedback-controlled systems through computational metaphors and systems of quantification from cybernetics and information theory. ((Edwards, The Closed World, 184.))
Integrations of human and machine on the level of neural transmission, storage and computation had clear applications in military surveillance and targeting systems. As Section Chief at the NIMH, Lilly sought to apply these ideas to the investigation and treatment of neurological disorders using electric brain stimulation as his material basis for what I call neuroelectric networks. Lilly conceptualized these networks as extensions of those within the brain (neural nets), and framed the neural network maps and brain-machine apparatuses produced in this work as a step forward in understanding the physiological mechanisms underlying not just sensory, motor and voluntary/spontaneous behavior but also the mechanisms by which these behaviors are learned. ((John C. Lilly, “Correlations between neurophysiological activity in the cortex and short-term behavior in the monkey,” in Biological and biochemical bases of behavior, ed. H.F. Harlow and C.N. Woolsey, (Madison: University of Wisconsin Press, 1958), 2. Learning, traditionally the domain of psychology rather than physiology, was the primary target of Lilly’s work. Lilly regarded his neuroelectric networks as part of a broader physiological “invasion” of psychology: “We come from the physiological side and intend [on] penetrating into the psychological territory … with physiological weapons.”)) Three characteristics of Lilly’s neuroelectric networks reflect the preoccupations of closed-world network-building more broadly: homeostasis, externalization, and mechanization.
In order to integrate human, animal and machine at scales both smaller (neural nets) and larger (global surveillance) than had ever been attempted before, network-builders like Lilly developed new inter-definitions human, animal and machine based on the concept of homeostasis, or the capacity of systems to maintain a steady state in unstable environments. At the same time as engineering, political doctrine and nuclear war strategy were constituting political struggle in terms of interlocked, homeostatic systems, cybernetic psychology attempted to reconceptualize the mind as a homeostatic system. ((Edwards, The closed world, 75-112. Though Edwards does not discuss homeostasis, the notion of “containment” operated in part as a metaphor for homeostasis in political and military discourse. The homeostatic view of the world as a set of self-contained systems was enacted through a series of gigantic “Big L” surveillance and communication projects developed by the U.S military, beginning with the Semi Automatic Ground Environment (SAGE), a radar and lookout system that monitored the skies for Soviet bombers.)) The most influential of these projects was McCullough-Pitts’ mathematical model of neural nets, wherein each neuron “fires” (transmits a chemical or electrical signal) in response to excitatory inputs that exceed a certain threshold. ((Lily E. Kay, “Cybernetics, information, life: The emergence of scriptural representations of heredity,” Configurations 5, no. 1 (1997): 65. In McCullough and Pitts neural net model of the brain, “Neurons were treated as black boxes, obeying mathematical rules governing the input and output of signals. A unit psychic event, or ‘psychon,’ was based on all-or-none (1,0: yes, no) impulses of neurons combined by convergence upon the next neuron to yield complexes of propositional events.” The influence of the McCullough-Pitts model of neural nets, and its potential application in the production of biological automata and the simulation of living systems, went well beyond neurology: “These logic-driven representations of biological mechanisms … served as epistemic models in the designs of military cybernetic systems of the 1940s and 1950s.” N. Katherine Hayles, How we became posthuman: Virtual bodies in cybernetics, literature and informatics (Chicago, IL: University of Chicago Press, 1999), 58. “The McCullough-Pitts neuron, as it came to be called, was enormously influential … for a generation of researchers it provided the standard model of neural functioning.”)) Neurons are connected by signal pathways to form networks, which are structured to enable logical propositions. In the McCullough-Pitts model, neural firing is the result of external stimuli relayed by sensory organs, and the structure of neural networks determines which sets of stimuli the mind calculates as “true” and “false.” In this way, neural networks redefine the mind as a homeostatic system by emphasizing the continuous loop between the external world and the neural “computation” of the world according to binary values of true and false.
Starting with the notion that the brain is already organized like a network (i.e., McCullough and Pitts’ neural nets), and that interconnections within this network continuously generate the mind, Lilly believed that detailed maps of activity within neural nets would help understand, diagnose and repair minds. By inserting a grid of electrodes into the brain, Lilly extended and externalized neural nets with electrical circuits and imaging technologies. The centrality of homeostasis to Lilly’s problematization of mind-machine networks was evident in the prominent role played by the circuit, both conceptually and materially. By expanding the neural portion of the loop through brain-implanted electrodes, Lilly’s circuits aimed to map and then alter the neural thresholds and connections and thus modify the material basis for thought.
By splicing the mind into extended neuroelectric circuits, Lilly’s project had the potential to articulate the mind to the then-controversial notion of reflexivity, which called into question the sharp division of mind and self from the external world. Lilly’s neuroelectric networks articulated instead a powerful vision of mind control through the application of homeostasis to externalized and mechanized neural networks. It is thus no surprise that military and intelligence organizations attempted to enlist Lilly in the effort to win the arms race in psychological warfare. If the mind could indeed be treated as a homeostatic system, regulating its relation with the environment through neural nets, Lilly’s neuroelectric networks promised to accelerate changes in thought and belief not through the long process of re-socialization whereby the subject internalizes new narratives about him or herself but by shuffling the subject between neuroelectric loops which determine the mind’s “steady state.”
Closed-world networks were designed to extract information from people, animals and machines. As Edwards demonstrates, the reorganization of the armed forces around nuclear weapons systems in the late 1950s led to the problem of how to monitor and coordinate these supersonic and devastating systems as they move through the ocean, the atmosphere and outer space. A key part of the solution to this urgent and complex coordination problem was the development of the “Big L” nuclear surveillance networks, which externalized the world in abstract, quantified models, and used those models to anticipate the actions of humans and machines.
A similar process of externalization characterized the networking of the “inner space” of the mind in closed-world discourse. Based on the cybernetic model of mind – the mind and its beliefs, memories, affects, perceptions and so forth as effects of the flow of control through neural circuits – the CIA and U.S. military initiated a covert research program in the 1950s and 1960s which aimed to “crack the code of human consciousness” through cybernetic psychology in order to develop new techniques of interrogation and persuasion. ((McCoy, A question of torture, 402.)) U.S. military and intelligence organizations invested billions of dollars into research aimed at developing techniques of “controlling human behavior.” ((CIA, as cited in Lemov, World as laboratory, 183.))
Whether Lilly had connections with the intelligence services before 1959 is uncertain but he explicitly aligned his sensory deprivation work at the NIMH with similar “psychic driving” research funded by the CIA at McGill University. ((John C. Lilly, “Some thoughts on brain-mind and on restraint and isolation of mentally healthy subjects” (presentation, Philadelphia Psychiatric Hospital, Philadelphia, November, 1955), 8. Lilly’s work at the NIMH on sensory deprivation was in turn cited by U.S. Navy medical researchers working in this area. See Thomas L. Myers, “Sensory and perceptual deprivation,” (presentation, Symposium on Medical Aspects of Stress in the Military Climate, Walter Reed Army Institute of Research, April 22, 1964), 2.)) The McGill experiments, conducted by Ewen Cameron, sought to “depattern” and “rewire” patients through a combination of psychotropic drugs, sensory deprivation and repetitive messaging, and effectively erased “the memories, personal habits, prejudices and neuroses, and even self-knowledge” of human test subjects. ((Lemov, World as laboratory, 207.)) Lilly’s neuroelectric networks at the NIMH attracted the attention of the military and intelligence establishment, primarily because his practices of electrical brain stimulation seemed to be capable of precise control over mental states, which would in turn facilitate interrogation and “learning” (or coercion). However, Lilly’s work aptly demonstrated that any project seeking to transfer information from machines into minds depended on the prior externalization of neural networks in the form of quantified values (thresholds) and “addresses” or point locations within the brain. If “Big L” networks were the culmination of the processes of standardizing space and time in the 18th and 19th centuries “so making different parts of the world locatable and transposable with a global architecture of address,” Lilly’s work aimed to externalize the “inner space” of the mind as a calculative and calculable space in which every mental state could be given a precise address in neuroelectric networks. To this end, Lilly reconfigured electrical brain stimulation techniques for a series of mapping expeditions in the brains of animal test subjects in the late 1950s. ((John C. Lilly, “Learning elicited by electrical stimulation of subcortical regions in the unanesthetized monkey,” Science, 125 (1957): 748. In these experiments, Lilly inserted electrodes into various regions of a monkey’s brain, sent electric current through the electrode, and recorded the animal’s behavior and its neurological activity. The implanted electrodes enabled Lilly to map neural networks and to link sensory events, muscle movements and other behaviors to patterns of activity in those networks.))
While most maps of transportation and communication networks represent the spatial structure of those networks, Lilly devised a method for representing neural transmissions over time in his network maps (Figure 1). The device consisted of a series of implanted electrodes which transmitted the relatively weak signals emitted by the brain through a cable to a set of amplifiers, which in turn triggered a grid of cathode glow tubes. Variations of light in the glow tubes were recorded by 16 mm motion picture camera running at 128 frames per second. The system enabled signals as short as 30 milliseconds to be visualized as a series changes in a grid of dots in a camera frame (Figure 2). The animal’s responses to external stimuli as well as its spontaneous actions appeared to contain distinct “figures” or “moving, spatially organized structures” of electrical signals. ((John C. Lilly, “Instantaneous relations between the activities of closely spaced zones on the cerebral cortex: Electrical figures during responses and spontaneous activity,” American Journal of Physiology 176 (1954): 496.)) In Lilly’s view, these experiments successfully produced a seamless map of neural networks. ((Lilly, “Correlations,” 9. Lilly states, rather triumphantly, “The maps of the movements elicitable from most of the lateral convexity of the hemisphere in an unanesthetized monkey with 610 implanted electrodes, have been completed … there are no gaps in the maps over the hemisphere.”)) The setup, including the visualization and recording techniques, constructed sensation, bodily movement and spontaneous activity as patterns of electrical activity occurring between interconnected neurons. ((See John C. Lilly, “Some evidence of gradients of motor function in the whole cerebral cortex of the unanesthetized monkey (abstract),” Proceedings of the 20th International Physiology Congress (1956), 567-8.))
For Lilly, cracking the code of neural communication required the externalization of the nervous language through the cataloguing of the brain’s sign system (signs were differentiated by waveform, amplitude, pulse sequence and duration) and mapping the neural circuits that send and store these signs. In order to externalize the structure and code of neural communication, Lilly neuroelectric networks bypassed the subject’s body through the development of new means of restraining the bodies of animal subjects in order to counter the subjects’ tendency (especially monkeys) to disconnect themselves from the network by pulling the electrical leads attached their brain implants (see Figure 3). ((John C. Lilly, “Development of a double-table-chair method of restraining monkeys for physiological and psychological research,” Journal of Applied Physiology 12 (1958): 134-6. Lilly devoted considerable time to the development of new types of restraint for the purpose of externalizing neural nets, and he presented these devices and methods at scholarly meetings and conferences and published short articles detailing these innovations. Lilly found that the restraints exerted a “taming influence” and accelerated the attachment of the animal subject to the researchers or personnel who fed them. In this way, the mapping project produced docile animal bodies and the techniques for producing such bodies enabled control over the subject’s disposition and attachments.)) A good network map for Lilly would be “stable and easily elicitable” in the live animal subject. ((Lilly, “Development,” 10.)) Lilly’s maps embody the closed-world preoccupation with re-engineering “suboptimal” systems; they not only record patterns of neural activity but also enable the mechanization of those patterns in live subjects.
The decision to establish nuclear weapons as the core of U.S. military force led to an intense concern with expanding the range, scope and speed of information transmission since, as Edwards explains, unlike conventional weapons, nuclear weapons “require instantaneous and massive responses, which must be preprogrammed because their execution must be virtually automatic.” ((Edwards, The Closed World, 131.)) Externalizing the world through new maps, metrics and simulations may have had intrinsic value to scientists seeking to understand the Earth, but with the deployment of nuclear weapons and “Big L” systems, externalizing practices were increasingly driven toward mechanization to reduce or even eliminate reliance on error-prone humans and thus provide the speed and accuracy of information demanded by nuclear weapons systems.
As Lilly began exploring, probing and reprogramming the minds of monkeys and dolphins in the 1950s, the material meanings of his interspecies networks became aligned with military and intelligence interests in what Jan De Vos calls techniques of “desubjectivization” or the effacement of models and senses of self. ((Jan De Vos, “Depsychologizing torture,” Critical Inquiry 37, no. 2 (2011): 311.)) The cybernetic model of the brain as a feedback-controlled system propelled Lilly’s attempt to erase and overwrite the programs governing the animal’s mind through neuroelectric and, later, bioacoustic networking. Lilly’s caution with regard to the implications of his work for brainwashing gave way in the early 1960s to a discourse of perfect control over thought and behavior. In one of his last papers on electrical brain stimulation, he suggested that
every bit of behavior and every bit of subjective life have their parallel in the electrical activity of the whole brain at each given instant … electrical currents suitably placed and timed within the brain can conceivably control the brain’s activities and hence the thoughts and emotions as well as the behavior of the whole organism. ((John C. Lilly, “The biological versus psychoanalytic dichotomy,” Bulletin of the Philadelphia Association for Psychoanalysis 11 (1961), 118.))
It was on the basis of this much bolder position on the electrical basis of thought and action that Lilly began to explore the possibility of using the neuroelectric network to “reprogram” vocalizing processes in dolphins in order to enable interspecies dialog. The first step towards reprogramming, however, was desubjectivization, or the erasure of programs that interfered with coercive persuasion.
The relationship between mapping and mechanizing neural networks was particularly evident in Lilly’s efforts at the NIMH to condition the behavior of animal subjects using the same apparatus that he used to map the neural network. Based on these experiments, Lilly began to describe the brain’s control mechanisms in cybernetic terms. ((See John C. Lilly, “Some considerations regarding basic mechanisms of positive and negative types of motivations,” American Journal of Psychiatry 115, no. 6 (1958), 498-504.)) Each region of the brain, he claimed, is interwoven with “start” and “stop” systems (or “punishing” and “rewarding” systems), which control neural activity. As Lilly and his team refined their stimulation and recording techniques, they found that they could accelerate behavior change by artificially stimulating motivational regions in the brain, most notably in Lilly’s (1957a) apparently successful attempt to train monkeys to use switches to indicate whether the electrodes were triggering “stop” or “start” systems. ((John C. Lilly, “Learning motivated by subcortical stimulation: The ‘start’ and ‘stop’ patterns of behavior,” in Reticular formation of the brain, ed. H.H. Jasper, L.D. Proctor, R.S. Knighton, W.C. Noshay and R.T. Costello (Boston: Little, Brown and Company, 1958), 78. “[T]he whole animal can be trained to give behaviorial signs of what goes on inside, in addition to the electrical and neuronal signs picked up inside the brain … the animal can give certain thresholds, which we eventually can relate to the electrical records.”))
Quantification and mapping of the neural network set the stage for the mechanization of communication in the brain. Lilly replicated the waveforms of neurological signaling with sequences of electrical pulses (or “stimulus trains”), which could be transmitted by a “moveable ‘roving’ electrode system” through the skull into the brain’s “stop” and “start” systems. ((Lilly, “Learning motivated by subcortical stimulation,” 79.)) In this way, affective states could be produced by means of brain implants rather than by external stimuli. According to Lilly, the pulse trains produced specific sets of behaviors in the test subjects. ((Lilly, “Learning elicited by electrical stimulation,” 2. Lilly found that the activation of “start” zones generated “mild ‘searching’ or ‘hallucinating’ behavior” and inhibited “attack or escape actions.” The same pulse trains produced very different behaviors in the “stop” zones, where “either pain-like or fear-like behavior can be reproduced in different regions with inhibition of feeding, grooming, and similar types of spontaneous activity.”))
While Lilly was initially quite careful not to suggest that this system allowed external “push-button control” of cognition or behavior, these experiments led to Lilly’s conceptualization of learning as a process of “feedback circuitry changes within cortex and between cortex and brainstem.” ((Lilly, “Learning motivated by subcortical stimulation,” 93.)) On the basis of this notion of learning as material changes in the “circuitry” or connectivity of neurological systems, Lilly quickly recognized that pulse trains delivered via brain implants to the brain’s “stop” and “start” zones, which instantaneously generate pleasurable or painful experiences, provided an effective means of negative and positive reinforcement in conditioning experiments. From a “powerful investigative tool” for mapping start and stop regions in neural network and linking events in that network to behavior, electrical brain stimulation became a conditioning system for altering neuronal connectivity and its behavioral “outputs.” ((Lilly, “Learning motivated by subcortical stimulation,” 78.))
Through the inter-definition of humans, animals and machines as components of homeostatic systems, the externalization of neural networks in quantifiable models, and the mechanization of neural communication to facilitate control over mental states, Lilly generated a series of conceptual and material overlaps between neuroelectric networks and the institutional matrix and techniques of psychological warfare. These overlaps were key to the stabilization of his networks in the 1950s and 1960s since they aligned neuroelectric techniques with the material meanings of networked media in the institutions that funded his work. Yet, as Lilly began conducting his experiments outside of his Bethesda NIMH facility, first at Marine Land, Florida and later at his own custom-built interspecies communication laboratory in St. Thomas, his network-building practices and the cultural logic underlying those practices began to shift away from the closed-world.
Drifting toward the Green-World
The overlap between Lilly’s neuroelectric networks and the military’s psychological warfare project could be seen as early as 1955 when Lilly tested the effects of electrical brain stimulation on Tursiops truncatus at Marine Studios near St. Augustine, Florida. ((See Burnett, Sound the whale and Wood, Marine Mammals.)) It was the same type of neuroelectric network that captivated his audience at a Pentagon meeting in 1959 with members of what Lilly called the “covert intelligence services,” who regarded the neuroelectric network as a means of mind control. ((John C. Lilly, The Scientist: A Metaphysical Autobiography (Berkeley, CA: Ronin Publishing, 1988), 87-97.)) Yet, from these interrogation-inspired experiments at Marine Studios, Lilly witnessed an unexpected distribution of communicative capacities. Since dolphins lack vocal cords, Lilly was rather startled in 1957 when, during a battery of neuroelectric tests, one of the dolphins mimicked the speech of the researchers. Upon activating the “start” system in the dolphin’s brain, the dolphin “told us about it by covering a large repertory of assorted complex whistles, Bronx cheers, and impolite noises.” ((John C. Lilly, “Some considerations regarding basic mechanisms of positive and negative types of motivations,” American Journal of Psychiatry 115, no. 6 (1958), 501.)) In this way, the effort to map the dolphin’s neural networks produced Lilly’s experience of dolphin vocalization, which in turn signified the dolphins’ “complex capabilities and great mental sensitivity.” ((John C. Lilly, “Productive and creative research with man and dolphin,” Archives of General Psychiatry 8 (1963), 112.)) Believing that he had witnessed novel behavior stemming from the plasticity of the dolphin’s extended neuroelectric network, Lilly left the NIMH to investigate the possibility that dolphins might possess an “advanced capability” for communicating beyond their own species. ((Lilly, “Productive and creative research with man and dolphin,” 112.))
The experience of “humanoid” dolphin vocalization in the neuroelectric network profoundly affected Lilly’s view of nonhuman communication and intelligence. In 1958, Lilly addressed the American Psychiatric Association at its conference in San Francisco and was expected to give a talk about the significance of his neuroelectric experimentation on animals for understanding and treating neurological disorders in humans. Lilly instead suggested that researchers ought to turn their attention to the problem of communicating with dolphins: “if we are to ever communicate with a non-human species of this planet, the dolphin is probably our best present gamble.” ((Lilly, “Some considerations,” 501.))
Later that year, he left the NIMH and established the Communication Research Institute, which abandoned electrical brain stimulation and the broader operant conditioning framework which informed Lilly’s work at the NIMH in favor of bioacoustic methods of bridging the gap between human and dolphin. Whereas the neuroelectric networks aimed to map and externalize the brain’s neural nets, bioacoustic networking focused attention on the interface between human and nonhuman biocomputers. This attentiveness to the human-nonhuman interface was evident in the construction of the “dolphin telephone,” in which a circuit of hydrophones converted the underwater acoustic exchanges between dolphins into airborne sound that could be heard and analyzed by the human researchers, as well as a portion of the St. Thomas laboratory called the “flooded house” wherein rooms, pools and balcony flooded with eighteen inches of water enabled “sonic reprogramming” of the dolphin through human-dolphin co-habitation. ((See John C. Lilly, The mind of the dolphin: A nonhuman intelligence (Garden City, NY: Doubleday, 1967).)) In the latter half of the decade, Lilly integrated these bioacoustic networks into what he called “studies and experiments in neuropsychopharmacology,” which refers to doses of LSD-25, acquired from sources at the NIMH, which Lilly administered to himself and to the dolphins in an effort to break the interspecies barrier. ((See John C. Lilly, Programming and metaprogramming in the human biocomputer(New York: Julian Press, 1968); Burnett, Sounding the whale, 616; Burnett, “Mind in the Water,” 49.))
Lilly’s increasingly eclectic network-building practices in the latter half of the 1960s alienated his sources of funding as well as many of his colleagues and supporters, some of whom began work on other U.S. Navy-funded projects on dolphin communication. What drove Lilly’s work in this period, and what drove many of his supporters away, was not so much the change in network materials (e.g., from brain implants to hydrophones and LSD) but rather the shift toward a cultural logic or organizational framework for network-building that would seem dangerous to those working within closed-worlds. Whereas closed-world networks are oriented toward the imposition of new forms of control on the world, Lilly, like the protagonists in Edwards’ green-world dramas, engage with the “problem of spiritual growth” by opening channels between humans and “mystical powers, animals, or other life forms.” ((Edwards, The Closed World, 312.)) Three key shifts in Lilly’s network-building practices articulated his move from closed-world to green-world.
First, the CRI’s bioacoustic networks can be understood as part of a broader shift in cybernetics from homeostasis to reflexivity as the key concept guiding the interdefinition of humans, animals and machines spliced into informational circuits. As N. Katherine Hayles argues, “Although the information feedback loop was initially linked with homeostasis, it quickly led to the more threatening and subversive idea of reflexivity,” that is, “the movement whereby that which has been used to generate a system is made, through a changed perspective, to become part of the system it generates.” ((Hayles, How we became posthuman, 8.)) The importance of reflexivity to Lilly’s project became clear in the “sonic reprogramming” experiments at the CRI wherein dolphins listened to recordings of human speech programs and were encouraged to adjust their own sound output programs accordingly (see Figure 4). ((See John C. Lilly, Alice M. Miller, and Henry M. Truby, “Reprogramming of the sonic output of the dolphin: Sonic burst count matching,” Journal of the Acoustical Society of America 43, no. 6 (1967), 1413.)) While these experiments began with the sharp division between human “programmer” and dolphin program, Lilly began to give detailed attention to the effects of new “programs” or beliefs, integrated into the human biocomputer through LSD, hypnosis and other means, on the researchers’ perception of the outcome of experiments. Audio recordings of dolphins’ efforts to mimic human speech were intended to function as documentation of the experiments but instead became media for reprogramming the beliefs of the human researchers. Rather than splicing animals into informational circuits and removing the scientist from the resulting closed-system, Lilly aimed to explore the capacity of a stipulated belief or program to become true to the mind and, as Jeff Pruchnic demonstrates, he empirically catalogued those beliefs acquired through metaprogramming as well as the “effects this endeavor produces on his/her current and subsequent behavior and sense of self.” ((Jeff Pruchnic, “Neurorhetorics: Cybernetics, psychotropics, and the materiality of persuasion,” Configurations 16, no. 2, 191.)) Cheerfully embracing the concept of reflexivity that many other postwar cyberneticians found deeply unsettling, Lilly effectively marked his networks as opposed to, and perhaps transcending, the closed-world.
Second, the shift from closed-world to green-world networks was articulated through the elaboration of ritualistic practices of incorporation, alongside the abstracting externalization techniques imported from the closed-world. The objective underlying Lilly’s neuroelectric networks at the NIMH was to inscribe or extract the thresholds and the topology of neural nets in quantified form, and these seemingly disembodied models of the brain’s informational circuits could then be used as diagrams for the mechanization of neural processes of learning and persuasion. Yet, as Hayles argues, from McCullough and Pitts onward, the concept of information as disembodied depended on a set of (embodied) “inscribing practices,” which increase the mobility of traces from the particularities of animate and inanimate bodies into generalized models, databases and simulations. It is not so much that the human-dolphin networks opposed inscription but that inscription became a means for the inward movement that Hayles terms “incorporating practices,” whereby action is “encoded into bodily memory by repeated performances until it becomes habitual.” ((Hayles, How we became posthuman, 199.)) The primary means of incorporation in Lilly’s work at the CRI was the process of “metaprogramming” the human biocomputer. For Lilly, metaprogramming designated the evolutionary shift in the biocomputers from haphazard learning to conscious and deliberate practices of “learning to learn.” Much of the CRI’s cultural practices aimed to probe the limits of this capacity in the context of interspecies communication. Lilly (1967) writes of the dolphin,
We have found that, in dealing with such a large-brained mammal, we must keep the working hypothesis in mind that “they are highly intelligent and are just as interested in communicating with us as we are with them” … If we use any other hypothesis, we have no success whatsoever in dealing communicatively with them. ((Lilly, Mind of the dolphin, 83.))
Experiments at the CRI began by stipulating a new belief and then inculcating this belief into the researchers until they behaved as if this belief were true. The researchers would then study the manner in which the enactment of the belief in experimental settings affected communication between humans and dolphins. Through repetitive bioacoustic and psychotropic techniques of metaprogramming, the belief generates new experiences (from the ecstatic to the traumatic) of human-dolphin communication, which are in turn incorporated into bodies since those experiences materially alter the neural networks of the observer and enable new kinds of interlocked actions between the observer and the observed.
Third, the transition to green-world networks was also marked by the development a renewed emphasis on real-time engagement with the world, alongside processes of mechanization (e.g., quantifying the dolphin’s neural networks and sensory organs in terms of bits per second). The CRI’s networks were directed not so much toward the mechanization of neural nets but rather toward real-time performance of software or “programs” stored in a “network of bodies.” ((John C. Lilly, The steersman: Metabeliefs and self-navigation (Berkeley CA: Ronin Publishing, 2006), 18.)) By creating, cataloguing, and altering programs in the network of bodies, the CRI aimed to generate new possibilities for interspecies communication that would leave a mark on the researchers’ neural structures and thus alter the system of belief governing subsequent performances, expectations, and recognitions of interspecies communication. For Lilly, interspecies communication research depended upon the discernment and revision of models of communication stored in human and dolphin biocomputers. This kind of research required a new system of “on-line” (real-time and continuously revised) knowledge production distributed across human and dolphin bodies:
When two huge computers (such as a dolphin and a human) interact in an on-line fashion the problem is not only the proving or the disproving of a hypothesis by new data acquisition. It is also the forcing of on-line testing continuously, and trying to elicit what model program of communication is stored currently in each of the computers. ((Lilly, Mind of the dolphin, 93-4.))
Conceptualized as a process of networking human and nonhuman biocomputers, technique, skill and expertise in interspecies communication shifted from information extraction and mechanization to navigation of the network of bodies through LSD, sensory deprivation and sonic persuasion. Like other skills that can only be demonstrated through performance, navigating the network of bodies had the potential for virtuosity, which was demonstrated at the CRI through, for example, Lilly’s increasingly prolonged sensory deprivation sessions in flotation tanks on doses of LSD (at least one of which culminated in hospitalization and a near-death experience) ((Cf. Burnett, The sounding of the whale, 612.)) and the seemingly interminable human-dolphin co-habitation experiments endured by Margaret Howe in the flooded house. The turn from the mechanization of neural nets to navigation through the network of bodies helped to place the CRI’s reputation within the scientific community in jeopardy. What is perhaps more significant here in terms of the shift from closed-world to green-world networks is that building interspecies communication networks based on a concept of embodied performance seemed to require the researchers to put their own bodies on the line, along with those of the dolphins.
Given the influence that military interests and techniques of surveillance and interrogation appear to have had on Lilly’s networks, and given that Lilly’s work had a significant impact on subsequent military-funded mind control and interspecies communication research programs, it is tempting to regard Lilly’s networks as symptoms of the broader weaponization of bioacoustics and brain science, and thus as an extension of the military and intelligence apparatus. The CRI’s human-dolphin networks, one might argue, did little more than produce knowledge for the elaboration of military networks of communication and control in minds and oceans.
This conclusion certainly captures important elements of the historical conditions of Lilly’s networks. However, this conclusion, which articulates the broader historical narrative of the expanding cage of calculative rationality in Cold War-era network-building, is liable to overlook significant shifts in Lilly’s practices, materials and discourses of interspecies networking, and the manner in which these shifts led to a growing tension between Lilly and his funding agencies. As I have argued, closed-world and green-world networks differ not so much in terms of their underlying technology but rather in terms of the cultural logic guiding the creation and maintenance of networks. By revisiting Lilly’s networks, I have shown that the neural nets were extended and reconfigured in radically different ways through drugs and circuitry. On the one hand, the dolphin’s neural nets could be treated as a closed system in order to facilitate the externalization and mechanization of neural activity. On the other hand, the same networking materials and techniques enabled Lilly to alter his own neural structures so as to incorporate new beliefs about the dolphin’s biocomputer and the programs stored therein, and to enact those beliefs in experimental practices in order to create further changes in neural structures and thus expand possibilities for unexpected behavior. The two cultural logics articulated the material meanings of networks in conflicting ways. Whereas Lilly’s closed-world networks sought to enclose and surmount the human/nonhuman divide for the purpose of establishing centralized command and control of minds on both sides of that divide, the green-world networks developed at the CRI aimed to transcend the human/nonhuman divide as part of a broader effort to restore a sense of cosmological unity.
Article: Creative Commons Attribution-ShareAlike 3.0 Unported License
Image: "Chapter 1: Paragraph 10”
From: "Drawings from A Thousand Plateaus"
Original Artist: Marc Ngui
Copyright: Marc Ngui