Neurotechnologies as weapons in national intelligence and defense – An overview James Giordano, PhD*1-3 and Rachel Wurzman, PhD(c)4

1. Center for Neurotechnology Studies, Potomac Institute for Policy Studies, Arlington, VA, 22203, USA, 2. Department of Electrical and Computational Engineering, University of New Mexico, Albuquerque, NM, 87131, USA, 3. Oxford Centre for Neuroethics, University of Oxford, UK, Email: [email protected], 4. Interdisciplinary Program in Neuroscience, Georgetown University Medical Center, Washington, DC, 20057, USA. Abstract Advances in neuroscience and neurotechnology have necessitated discussions on the ways that such developments could be used as weapons in contexts of national security, intelligence, and defense. This paper defi nes the concept of neuroweapons, and elucidates operational issues associated with their use to aid informational and strategic intelligence, such as brain-machine interfaces to improve effi ciency in data analysis. As well, exploration of neuropharmacologic, neuromicrobiological, and neurotoxic agents are discussed relevant to their utility in combat scenarios. The limitations of emerging neurotechnologies as weapons are addressed, as both regards practical and operational frameworks, and implications relevant to formulation of ethico-legal guidelines and governance of research, development and potential use.

Key words: neuroscience, neurotechnology, neuroweapons, neurosecurity, national defense, biotechnology, weaponry

Introduction summarized the state of neuroscience as relevant to the 1) potential utility for defense and intelligence applica- Advances in neuroscience have progressed rapidly over tions, 2) pace of progress, 3) present limitations, and 4) the last two decades. The fi eld has become increasingly threat value of such science. In characterizing the chal- interdisciplinary, and has been a nexus for the develop- lenges to advancing defense-oriented neurotechnologies ment and use of a wide range of technological innova- — as well as maintaining the United States’ international tions (viz.- neurotechnology). While usually considered competitive edge — the committee noted that a signifi cant in medical contexts, many neurotechnologies may also problem was the “…amount of pseudoscientifi c informa- be viably engaged as weapons. Such “neuroweapons” tion and journalistic oversimplifi cation related to cogni- are obviously of great interest in and to national security, tive neuroscience.” (1) intelligence and defense (NSID) endeavors, given both the substantial threat that these technologies pose to the Given the relative nascence of neuroscience and much defense integrity of the US and its allies, and the viability of neurotechnology, the development and use of neu- of these approaches in the US NSID armamentarium. A roweapons discussed in this essay are incipient, and in 2008 report entitled Emerging Cognitive Neuroscience some cases, the potential utility of these approaches is and Related Technologies by the ad-hoc Committee on speculative. But any such speculation must acknowledge Military and Intelligence Methodology for Emergent that neurotechnological progress in these areas is real, Neurophysiological and Cognitive/Neural Science and therefore consideration of the potential trajectories Research in the Next Two Decades (National Research that neurotechnologies-as-weapons might assume is both Council of the National Academy of Sciences (hereaf- important and necessary. As well, such discussion must ter referred to as the 2008 NAS ad-hoc committee) (1) entail a pragmatic view of the capabilities and limitations

Synesis: A Journal of Science, Technology, Ethics, and Policy 2011 T:55 © 2011 Potomac Institute Press, All rights reserved Synesis: A Journal of Science, Technology, Ethics, and Policy 2011 of these devices and techniques, and the potential pitfalls (e.g., prolonged fl ashing lights, irritating music, and sleep of — and caveats to — their use. Herein, we address deprivation to decrease resistance to interrogation) (5). 1) the possible ways that neurotechnologies can be utilized Even the distribution of emotionally-provocative propa- as weapons; 2) the NSID aims that might be advanced ganda as psychological warfare could be considered to be by neuroweapons; and 3) some of the consequences and/ an indirect form of neuroweapon (8). or implications of developing neurotechnologies toward these ends. While such an expansive consideration may be important to evaluate the historicity, operational utility, and practical What is a neuroweapon? (and ethical) implications of neurotechnology-as-weapons, in this essay, we seek to provide a concise overview of neu- A weapon is defi ned as “a means of contending against roweapons, and therefore restrict discussion to applications another “ and “…something used to injure, defeat, or de- of emergent technologies of cognitive neuroscience, com- stroy” (2). Both defi nitions apply to neurotechnologies putational neuroscience, neuropharmacology, neuromicro- used as weapons in intelligence and/or defense scenarios. biology, and neurotoxicology. The former approaches (e.g., Neurotechnology can support intelligence activities by cognitive and computational neuroscience; neuropharma- targeting information and technology infrastructures, to cology) could be used for more indirect (yet still neurocen- either enhance or deter accurate intelligence assessment, tric) applications, including the enablement and/or enhanc- the ability to effi ciently handle amassed, complex data, ing of human efforts by simulating brain functions, and the and human tactical or strategic efforts. The objectives for classifi cation and detection of human cognitive, emotional neuroweapons in a traditional defense context (e.g., com- and motivational states to augment intelligence, counter- bat) may be achieved by altering (i.e., either augmenting intelligence, or human terrain deployment strategies. The or degrading) functions of the nervous system, so as to af- latter methods (e.g., neuromicrobiology, neurotoxicology, fect cognitive, emotional and/or motor activity and capa- as well as neuropharmacology) have potential utility in bility (e.g., perception, judgment, morale, pain tolerance, more combat-related or special operations’ scenarios. or physical abilities and stamina). Many technologies (e.g., neurotropic drugs; interventional neurostimulatory Contending against potential enemies: Neuro- devices) can be employed to produce these effects. technology within information infrastructures and intelligence strategy As implements that target, measure, interact with, or sim- ulate nervous system function and processes, the use of Those neurotechnologies that enhance the capabilities of neurotechnololgies as weapons are by no means a new the intelligence community may be considered weapons innovation, per se. Historically, such weapons have in- in that they provide “…a means of contending against an cluded nerve gas, and various drugs. Weaponized gas other” (2). Certain neurotechnologies may be particularly has taken several forms: lachrymatory agents (a.k.a., tear well suited to affect performance in, and of the intelli- gases), toxic irritants (e.g., phosgene, chlorine), vesicants gence community. The tasks of both human analysts and (blistering agents; e.g., mustard gas), and paralytics (e.g., the technologies they use are becoming evermore recipro- sarin). These may seem crude when compared to the ca- cal and inter-dependent. Without technology to pre-pro- pabilities of the more sophisticated approaches that can cess and sort large quantities of complicated information, be used today — or in the near future (3). Pharmacolol- human analysts could not obtain a cohesive picture from gic (e.g., amphetamines) and various ergogen- which to draw necessary inferences about the capabilities ics (e.g., anabolic steroids) have been used to augment and intentions of (friendly, neutral or hostile) intelligence combatant vigilance, and sedatives (e.g., ) targets. Yet, information technology presently requires have been employed to alter cognitive inhibition and fa- human programming and implementation of human-con- cilitate cooperation during interrogation (3-7). Sensory ceived models to parse the volume and types of informa- stimuli have been applied as neuroweapons: some to di- tion collected. Moreover, some information remains prob- rectly transmit excessively intense amounts of energy to lematic to collect (e.g., attitudes and intentions of human be transduced by a sensory modality (e.g., sonic weap- subjects). Neurotechnologies that would facilitate and en- onry to incapacitate the enemy), while others cause harm hance these capabilities might decrease the fallibility of by exceeding the thresholds and limits of tolerable ex- “human weak links” in the intelligence chain through the perience by acting at the level of conscious perception application of neurally-yoked, advanced computational

T:56 Synesis: A Journal of Science, Technology, Ethics, and Policy 2011 strategies (i.e., brain-machine, and machine-brain inter- mittee identifi ed such technology as a potential threat, but faces; BMI/MBI respectively) in the management and one that remains largely theoretical — at least at present integration of massed data. Similarly, neurotechnologies (1). Such computational cognitive frameworks may “bor- can be developed to manage the increasingly signifi cant row” human capabilities, not by mimicking processes in problem of the sheer volume of cyber-based communica- the brain (which may not be suffi ciently well understood tions that has threatened intelligence systems with inunda- to begin with), but by involving conceptual components tion. The widespread and inexpensive use of sophisticated of idealized neurally-modeled systems that are linked in communication technology (e.g., social media), and dif- ways that enable performance of similar — if not more fi culty of allocating resources to gather intelligence-focal rapid and advanced — neuro-cognitive functions. More- “signals” over evermore increasing, non-relevant “noise” over, neurally-coupled hybrid systems could be developed has made more coherent collection and interpretation of that link computational interfaces to human neuronal activ- intelligence information a priority (1, 9). ity, so as to optimize Bayesian-like predispositions to cer- tain types of stimuli (18). This would limit input datasets The principal neurotechnologies that can be used in these to more critical features, and thereby allow more effi cient tasks are distributed human-machine systems that are (i.e., rapid and accurate) detection, observation, orientation either employed singularly, or linked to networked hier- (and decisions) by the human user. Conjoinment and reci- archies of sophisticated BMIs, to mediate access to, and procity could be used to enhance the feature-detection and manipulation of signal detection, processing and/or inte- intelligence capacities of both (the machine and human) gration. Neurotechnologic innovations that are capable of systems. processing high volume, complex datasets are forms of physiomimetic computing hardware (1). Such hardware Enhancement of neural and cognitive capabilities may leverages analog, rather than digital components, with “a be further achieved through some form(s) of cybernetics, continuous set of values and a complex set of connec- broadly considered as a feed-back and feed-forward sys- tions,” based on an understanding of neural networks as tem that obtains iterative re-assessment and modifi cation more than mere binary switches. An analog circuit ap- capacities through ongoing interactions between an agent proach would address current “modeling and simulation and its environments (19). According to the classifi ca- challenges”, be smaller and “…easy for the US — and its tion scheme of Clynes and Kline (20), the use of human- adversaries — to construct” (1). As well, given the analog machine interfaces can be regarded as a level 2 or level 3 nature of the magnetic fi elds used for real-time comput- cybernetic organism (viz.- a cyborg) in that it entails both ing, a small, portable, physiomimetic computer of this natural and artifi cial systems that are functional, portable, type might be uniquely valuable for applications of high- and/or biologically integrated. Cybernetic and cyborg density information processing (10-17). systems can be seen as sophisticated distributed human- machine networks, such as integrated software or robotic Information systems could conceivably be conjoined so that augmentations to human-controlled activity, that would neural mechanisms for assigning and/or detecting salience fuse and coordinate the distinct cognitive advantages of (i.e., processes involving cortical and limbic networks) may humans and computers. As stated in the NAS ad-hoc com- be either augmented or modeled into neurotechnologic de- mittee report, these systems could assist “…advanced vices for rapid and accurate detection of valid (i.e., signal sensory grids, control and could control unmanned au- vs. noise) information within visual (e.g., fi eld sensor, sat- tonomous systems, advanced command posts and intel- ellite and UAV-obtained images) and/or auditory aspects ligence analyst workbenches, coordination of joint or (e.g., narratives, codes) of human (HUMINT) or signal coalition operations, logistics, and information assur- intelligence (SIGINT). Formulating and testing credible ance” with consequences that “enhance the cognitive or hypotheses while monitoring large amounts of information physical performance of war fi ghters and decision-mak- could be accomplished by computational cognitive frame- ers or allow them to coordinate the actions of autono- works that are capable of both self-instruction (e.g., using mous systems with much-improved effectiveness” (1). the internet as a “training environment”), and learning from These systems would be of evident utility to multiple experience (e.g., via direct access to the operational envi- forms of intelligence acquisition and processing at both ronment). This articulates a form of artifi cial intelligence tactical and strategic levels. (AI) that functions to model — and embellish — human neural systems in cognition. The 2008 NAS ad-hoc com-

T:57 Synesis: A Journal of Science, Technology, Ethics, and Policy 2011

Strategic intelligence neuro- and psychologically viable approaches, there is in- terest in neurotechnology to augment the role, capability, Strategic intelligence is defi ned as gathering and analyz- and effect(s) of psychological operations as a “force mul- ing information regarding the capacities and intentions of tiplier” in both political and military tactics. This trend foreign countries and actors; it may also encompass po- began with the 1985 Department of Defense (DoD) psy- litical intelligence, given that “…[political intelligence] is chological operations’ (PSYOP) master plan and has been at once the most sought-after and the least reliable of the accelerated by the challenges posed by insurgencies in the various types of intelligence. Because no one can predict present confl icts in Iraq and Afghanistan (21). with absolute certainty the effects of the political forces in a foreign country, analysts are reduced to making fore- Such challenges emphasize the problems of cultural intel- casts of alternatives based on what is known about politi- ligence, and how these generate psychosocial obstacles cal trends and patterns” (9). The complex dynamics of po- to achieving tactical ends in the region. Tactical defi cits litical forces that contribute to such predictive diffi culty may be related to the military approach to psychological- are due, in part, to the numerous and varied agents in- political warfare as being centered upon a “confl ict of volved, all of whose actions are individually determined. ideas, ideologies, and opinions” while not adequately em- Thus, understanding the bio-psychosocial factors that in- phasizing notions such as “cultural and political symbols, fl uence individual and group dynamics, and being able to perceptions and emotions, behavior of individuals and detect these variables with high ecological validity (i.e., groups under stress, and cohesion of organizations and “in the fi eld”, under real-world conditions) is important to alliances” (22). Even if we were aware of such variables, both descriptive/analytic and predictive intelligence ap- we might still be fl ummoxed in infl uencing “the minds proaches. and hearts” of enemy combatants, because of the failure to correctly defi ne and predict which factors may affect A combination of 1) advanced socio-cultural neurosci- aspects of psychological warfare (such as the severance entifi c models of individual-group dynamics based upon or formation of alliances and collectives’ reactions to the theories of complexity adapted for use(s) in anthropology; threat of integrity). 2) suffi cient computing and BMI frameworks (perhaps as speculated above), and 3) certain forms of neuroimaging Thus, an appeal of neurotechnology is its (theoretical) po- and magnetoencephalography to accurately detect the tential for use in 1) defi ning substrates and mechanisms brain states and decision-biases of key or representative operative in culturally-relevant cognitions and behaviors, individuals might enable dramatically improved forecast- and 2) directly affecting perceptions, emotions, behaviors, ing of behavior patterns that are infl uential to socio-po- and affi liative tendencies. The most obvious possibility is litical change. These forecasts could include the descrip- the use of neurotechnology to assess and affect cognitive tion of neuro-cognitive states of specifi c agents/actors, capability, mood and/or motivation. Various forms of neu- the propagation dynamics of an idea or cultural construct, roimaging have been considered, as have the concomitant and/or node-edge cognitive and behavioral interactions of use of neurogenetics and neuroproteomic approaches in individuals and cohorts- any and all of which might be this regard. However, cognitive and emotional effects viable to identify specifi c targets for subsequent manipu- in individuals and across a population are complicated, lation (via other neuroweapons). and can often be unpredictable. Hence, a main criticism of neuroimaging is that although relatively valid and reli- However, intentions, as opposed to corresponding cogni- able in assessing individual mechanisms and substrates of tive and/or emotional states and their associated neuronal cognition and emotion under controlled (i.e., experimen- signatures, are diffi cult to detect using existing neurotech- tal) situations, the ecological validity of such protocols is nologies. This affects and alters the modeling approaches questionable, and thus neuroimaging may be of limited that could — and should — be used to describe or predict value in depicting more subtle cognitive-emotional, and individual or group activities. As well, it is important to motivational states, such as deception in “real-world” sce- consider the potential of technological interventions to al- narios (23-24). Adding to this is that neuroimaging is not ter events. Here, lessons may be garnered from experience a subtle technique, and protocols for assessing cognitive- with psychological warfare (6). Sometimes, techniques emotional variables would need to be explicitly concerned and tactics will induce unintended, if not frankly contrary with the ways that the testing environment affects individ- effects and results. Given the overarching applications of uals being evaluated. Neurogenetics and neuroproteomics

T:58 Synesis: A Journal of Science, Technology, Ethics, and Policy 2011 could enable assessment of predispositional variables and ylphenidate (28), and wakefulness-promoting agents (eu- even phenotypic characteristics that infl uence cognitions, geroics) such as the novel dopaminergic reuptake blocker emotions and behaviors, but these approaches are of only and histamine and orexin potentiating agent modafi nil limited predictive value, given the non-linear relationship (29); 2) somnolent agents such as the barbiturates, ben- of genetics to both phenotype(s) and the ultimate expres- zodiazepines and certain opiates (30); 3) mood altering sion of cognitive states and behavioral actions (25). Of agents such as the azaspirone (e.g., buspirone; course, today’s limitations often represent the challenges (30)), beta-adrenergic antagonists (e.g., propranolol, that and opportunities for tomorrow’s technology, and ongo- has been considered for its effects in decreasing agitation ing work is dedicated to use of a more convergent scien- and anxiety associated with traumatic events (30)), as well tifi c and technological paradigm to compensate for extant as dopamine and serotonin agonists (that at higher doses constraints and limitations, and create technologies that have been shown to induce fear, and psychotic symptoms are effective and easily employed/deployed in operational including paranoia (31)); 4) “affi liative” agents such as the settings (26). neurohormone oxytocin (32), and the substituted amphet- amines (e.g., methylenedioxy , MDMA Neuroweapons in combat scenarios — “ecstasy” (33); and 5) , such as acetylcho- line-agonists and gamma aminobutyric acid (GABA) re- A considerably more imposing possibility is to “change ceptor antagonists (34). The actions and effects of these minds and hearts” by altering the will or capacity to fi ght categories of drugs are provided in Table 1. through the use of neuropharmacologic, neuromicrobio- logical and/or neurotoxic agents that 1) mitigate aggres- While some of these agents can be used to enhance the sion and foster cognitions and emotions of affi liation or neuro-cognitive and motor performance of (one’s own) passivity; 2) incur morbidity, disability or suffering and troops (e.g., low does of stimulants, mood altering drugs, in this way “neutralize” potential opponents, or 3) induce etc), others have apparent utility against hostile forces mortality. James Hughes (27) has identifi ed 6 domains (e.g., somnolent, psychotogenic, affi liative, and convul- of neurocognitive function that can currently be phar- sant agents). Moreover, while a “weapon” is characteristi- macologically manipulated; these are 1) memory, learn- cally considered to be a tool used to incur injury, agents ing, cognitive speed; 2) alertness and impulse control; 3) such as oxytocin and/or MDMA may actually reduce or mood, anxiety, and self-perception; 4) creativity; 5) trust, prevent harm infl icted on an opponent by decreasing their empathy, and decision-making; and 6) waking and sleep- desire to fi ght or making them more amenable to affi lia- ing. As well, movement and performance measures (e.g., tion. These effects are wholly consistent with the more speed, strength, stamina, motor learning, etc.), could also formal defi nition of a weapon, as “…a means of contend- be enhanced or degraded (9). ing against another” (2). To paraphrase Kautilya, the per- son who becomes a friend is no longer an enemy (35). Neurotropic drugs Yet, this too can be viewed as potentially harmful in that drug-induced effects upon cognition and emotion may al- As mentioned previously, the use of neuropharmacological ter the identity, autonomy, and free will of others, and in agents to affect cognitive, emotional and behavioral so doing, are exercises of “biopower” (36-38). Neverthe- abilities is certainly not novel (vide supra). However, an less, we opine that when attempting to balance benefi ts, expanded capital of neuroscientifi c knowledge, namely the risks and harms within contexts of jus ad bello and jus increased understanding of molecular and systems-based, in bellum, such outcomes—while powerful—may need to structure-function relationships of the brain, has fortifi ed be considered as less injurious than either more profound depiction of substrates and mechanisms that are viable forms of neuro-psychiatric insult, or those produced by pharmacologic targets. Such knowledge, when coupled to more “traditional” weaponry. advancements in pharmaceutical technology, has allowed discovery and development of neurotropic agents with To be sure, the use of drugs to affect cognitive, emo- greater specifi city, potency, and bioavailabilty. tional or behavioral function carries many potential risks of abuse (e.g., excessive doses or too-frequent In general, drugs that have utility in combat and/or special use), misuse, unintended interactions with other drugs, operational settings include 1) cognitive and motoric stim- foods and situations, and alterations in social behavior ulants such as the chain-substituted amphetamine, meth- (27). Additionally, the effects of any drug depend on

T:59 Synesis: A Journal of Science, Technology, Ethics, and Policy 2011

Table 1. Neuropharmacologic Agents

Category Type/Drugs Principal Actions/Effects Side effects/very high dose effects Cognitive/motoric CNS Stimulants Increase DA/NE turnover/release Loss of appetite, stimulants (e.g., amphetamines1, Increase arousal insomnia, dizziness, methylphenidate1, pemoline2) Increase attention agitation, increased heart Elevate mood rate, dry mouth, high- Induce rebound depression and anxiety frequency tremor, or restlessness (e.g., modafi nil, Increase DA turnover Excitation or agitation, adrafi nil) Decrease DA reuptake insomnia, anxiety, Elevates hypothalamic histamine levels irritability, nervousness, Potentiate action of orexin aggressiveness, confusion, Wakefulness & decreased fatigue tremor, palpitations, sleep Increase arousal disturbances, or nausea Increase attention Few autonomic side effects Non- cognitive Potentiate AMPA receptor-mediated Possible headache, enhancers (ampakines6 neurotransmission somnolence, nausea, e.g., ampalex, farampator, Increase attention or impaired episodic phenotropil) Increase alertness memory (Farampator) No PNS stimulation Enhance learning and memory Increase tolerance to cold and stress Other Nootropics (racetams6 Potentiate muscarinic ACh receptor activity CNS stimulation, e.g., piracetam, , Activate NMDA/glutamate co-localized excitation, depression, aniracetam) with ACh receptors dizziness, sleep Non-specifi c increase in neuronal disturbances excitability (via AMPA and NMDA-receptor potentiation/ activation) Enhance learning and memory (nootropic) Anti-emetics Anti-spasmodics Monoamine reuptake DA/NE reuptake inhibitors Dry mouth, blurred inhibitors1,2 (e.g., buproprion2, Antidepressant effects vision, and dizziness, atomoxetine, , Decrease anxiety Buproprion causes venlafaxine, mirtazapine ) Increase concentration seizures at high doses Solmnolent and Benzodiazepines3 (e.g., Increase GABA binding Blurred vision, headache, tranquilizing lorazepam, prazepam, Increase neural inhibition confusion, depression, agents clonazepam, oxazepam, Increase somnolence impaired coordination, , midalzolam, Decrease arousal trembling, weakness, alprazolam) Decrease reaction time/coordination memory loss Motoric lethargy Anterograde amnesic effects 3 Azaspirones (e.g., buspirone, 5HT1A receptor agonists Dizziness, gepirone) Decrease NE activity nausea, headache, Decrease arousal nervousness, insomnia, Decrease agitation lightheadedness Decrease anxiety Adrenergics3 (e.g., clonidine, Stimulate α-adrenergic receptors Dry mouth, dizziness, guanfacine) NE autoreceptor agonist constipation Sedative effects Reduce heart rate Relax blood vessels

T:60 Synesis: A Journal of Science, Technology, Ethics, and Policy 2011

Table 1. Neuropharmacologic Agents, Continued

Category Type/Drugs Principal Actions/Effects Side effects/very high dose effects Solmnolent and Barbiturates (e.g., Potential for lethal Bind to α subunit of GABAA receptors, tranquilizing phenobarbitol, mephobarbitol, Inhibit AMPA receptor-mediated glutamate overdose via respiratory agents (cont’d) thiopental, amobarbital, activity depression secobarbital, pentobarbital) High doses may inhibit neurotransmitter release Block ligand-gated cation channel receptors

(nicotinic, glycine, 5-HT3 receptors) Decreased arousal Decreased concentration Impaired coordination Slurred speech Decreased anxiety Muscle relaxants (e.g., CNS-acting muscle relaxers Potential anti- carisprodol, cyclobenzaprine, Anti-cholinergic cholinergic effects, dry metaxalone) Potentiation of NE mouth, blurred vision,

5HT2 antagonist; decrease descending 5-HT constipation, memory activity loss effects Increases solmnolence

Imidazopirydine hypnotics Potentiate GABAA receptors Hallucinations, sleep (e.g., , ) Increase slow-wave sleep walking Facilitate sleep initiation (not maintenance)

Antipsychotics/ Neuroleptics Block D2 DA receptors/ autoreceptors Sedation, blood pressure (dopamine antagonists e.g., Decrease DA release or heart rate changes, chlorpromazine, haloperidol, Decrease agitation or dizziness, cognitive fl uphenazine, thioridazine, Increase sedation dulling loxapine, thiothixene, pimozide) Hypothalamic effects on metabolism, body temperature, alertness, muscle tone, and hormone production 4 Atypical antipsychotics Block D2 receptor types in limbic-system Tardive dyskenesia (e.g., clozapine, risperidone, Decrease dopamine release (uncontrollable writhing olanzapine) Fewer extrapyramidal effects movements), neuroleptic Some are anti-cholinergic (clozapine, malignant syndrome perphenazine) (hyperpyrexia)

May block 5-HT2A , α1 epinephrine, and H1 histamine receptors Anticonvulsants3 (e.g., Solmnolence, gabapentin, pregablin) hypertonicity, abnormal Analgesic in neuralgia gait, coordination, or CNS depressants movements, vision problems, confusion, eurphoria, and/or vivid dreams

T:61 Synesis: A Journal of Science, Technology, Ethics, and Policy 2011

Table 1. Neuropharmacologic Agents, Continued

Category Type/Drugs Principal Actions/Effects Side effects/very high dose effects Mood-altering Other monoamine antagonists5 Prevent DA, NE, and 5-HT transport into Nausea and vomiting, agents (e.g., reserpine, tetrabenazine) synaptic release vescicles severe depression, Deplete DA in presynaptic terminals drowsiness, dizziness, Depress mood and motor activity and nightmares Akasthesia, restlessness Beta-blockers5 (e.g., β-adrenergic antagonists Dizziness/light- propranolol) Decrease autonomic stress response headedness, drowsiness, Decrease performance anxiety insomnia/dysomnia Reduce posttraumatic agitation and anxiety Dissociatives (e.g., NMDA antagonists Muscle weakness, ataxia, : PCP, ketamine) Alters monoamine release/reuptake and loss of consciousness Inhibit DA release in frontal cortex through (ketamine), self- presynaptic NMDA receptors (PCP) destructive or self- Increase DA release/ inhibit DA reuptake in injurious behavior, limbic areas (PCP) impaired judgment Dissociative anesthesia (marked by catalepsy, amnesia, and analgesia) Induce/exacerbate psychosis (with euphoria/ dysphoria, paranoia, delirium, multisensory hallucinations) Ketamine induces anesthesia via σ- and μ- opioid receptors

Psychedelics (tryptamine 5-HT2A agonists Dry mouth, nausea, and alkaloids e.g., ibogaine, Decrease 5-HT reuptake vomiting , psilocybin, LSD) Loss of coordination Psychedelic/hallucinogenic effects

σ2 receptor, nicotinic receptor, NMDA antagonism Dopamine agonists (e.g., Inhibit DA reuptake/ breakdown, Induce nausea and L-DOPA, tyramine, pargyline, Increase DA release, or vomiting benztropine, apomorphine* Act as DA receptor agonists* Cause abnormal ropinerole*, bromocriptine*) May induce psychosis movements Increase compulsive behavior Enhance paranoia and/or fear response Affi liative agents Amphetamine derivatives Increase net release of monoamine Thermal dysregulation, (e.g., 3,4-methylenedioxy- neurotransmitters (5-HT & NE > DA) muscle tension, methamphetamine, MDMA Decrease 5-HT reuptake headache, dry mouth, a.k.a., “ecstasy”) Increased wakefulness, endurance; over-arousal (fl ight postponement of fatigue and sleepiness of ideas, insomnia, Sense of euphoria and heightened well-being agitation, hyperactivity), Sharpened sensory perception panic attacks, delirium, Greater sociability and extroversion depersonalization, mild Increased tolerance and desire for closeness hallucinations, or brief to other people psychosis is possible, neurotoxic with chronic exposure Oxytocin6 Acts at CNS oxytocin receptors Pro-social feelings, Evokes feelings of contentment, calmness, may sometimes and security include ethnocentrism, Reduces anxiety territoriality, and Increases trust and bonding xenophobia Decreases fear Stimulates uterine contractions in pregnancy

T:62 Synesis: A Journal of Science, Technology, Ethics, and Policy 2011

Table 1. Neuropharmacologic Agents, Continued

Category Type/Drugs Principal Actions/Effects Side effects/very high dose effects

Epileptogenics Inverse GABAA Act at benzodiadepine-binding site of receptor agonists GABAA receptors (non-volitile e.g., DMCM, Decrease GABA-binding affi nity of GABAA ; volitile e.g., –receptors fl urothyl6) Decrease frequency of chloride channel opening (to decrease inhibition) , stimulant, and effects Inhalation of fl urothyl elicits seizures mechanism not well-understood

Competitive GABAA Bind competitively to GABAA receptors at Convulsant effects antagonists (e.g., , GABA-binding site bucuculline) Reduce synaptic inhibition of neurons by decreasing chloride conductance

Non-competitive GABAA PTZ also increases neuronal excitability by High doses cause antagonists (e.g., affecting cation currents via NMDA, 5-HT1A, convulsions bergamide, pentetrazol a.k.a., 5-HT3, glycine receptors PTZ) Anxiogenic Circulatory and respiratory stimulant Glycine Antagonists (e.g., antagonist Highly toxic, , ) ACh receptor antagonist convulsions, death by Initial effects of nervousness, restlessness, asphixiation muscle twitching, neck stiffness give way to pupil dilation and convulsions of increasing severity Mixed GABA antagonist/ Positive of AMPA receptors Seizures at higher doses glutamate agonist (e.g., Inhibit desensitization of AMPA receptors ) Potentiate glutamate currents

Negative allosteric modulator of GABAA receptors Convulsant without neurotoxic effects Ionotropic glutamate receptor Enhances glutamate effects through kainate receptors Neuronal damage agonists (e.g., ) CNS stimulant Excitotoxic convulsant Muscarinic agonists (e.g., Non-selective muscarinic ACh receptor Excessive sweating and pilocarpine) agonist salivation, bronchospasm Systemic injection leads to chronic seizures and increased bronchial mucus secretion, vasodilation and bradycardia Non-specifi c cholinergic agonist Acetylcholinesterase inhibitor Convulsant at high doses (e.g., physostigmine) Increases synaptic ACh levels Indirectly stimulates nicotinic and muscarinic receptors Local anesthetics (e.g., Block fast voltage gated sodium channels With higher doses, lidocaine, prilocaine) Inhibit neuronal fi ring drowsiness, loss Moderate systemic exposure causes CNS of consciousness, excitation, symptoms of nervousness, respiratory depression dizziness, blurred vision, tinnitus and tremor and complete apnea Seizures followed by depression

1 also elevates mood or antidepressant effects; potential as mood-altering agent(s) 2 also lowers seizure thresholds; potential epileptogenic effects, especially at high doses 3 also anxiolytic; potential mood-altering agent(s) 4 also decreases agitation/psychosis and/or antidepressant effects; potential as mood-altering agent(s) 5 signifi cant sedation side effects; potential as solmnolentic agent(s) 6 mechanism not fully understood or effects inconclusively established at this time Note: ACh: acetylcholine; CNS: central nervous system; DA: dopamine; GABA: gamma amino butyric acid; 5-HT: 5-hydroxytryptamine (serotonin); NE: norepinephrine; NMDA: n-methyl D-aspartate

T:63 Synesis: A Journal of Science, Technology, Ethics, and Policy 2011 an individual’s particular combination of genes, envi- Here too, use of nanotechnology to preserve stable forms ronment, phenotype, and the presence or absence of both of pathogenic agents could be important for producing physiological and psychopathological traits, and these more durable aerosolized neurobioweapons (43). While can vary widely within a given population. Despite the these agents are capable of inducing large-scale infections relatively small size of the military, considerable diversity in a given population, such mass casualty effects may not still exists in the aforementioned characteristics, and this be required or desired. Instead, using these agents in more would need to be accounted for, as would any variations punctate approaches might be of greater advantage. Such in those populations in which the use of neurotropic agents techniques include: 1) inducing a small number of rep- is being considered. resentative cases (with high morbidity and/or mortality) that could incur mass public reaction (e.g., panic and/or Thus, it is probable that any neurotropic agent would paranoia) and impact upon public health resources – in- produce variable responses and effect(s) in a population clusive of a strained public-governmental fi duciary; 2) refl ective of individual geno- and phenotypes, as well as targeting specifi c combatants to incur health-related ef- biological variation in given individuals over time. This fects upon operational infrastructures; or 3) “in close” could incur an increased likelihood of unanticipated ef- scenarios in which particular individuals are targeted for fects; therefore, it is important that pharmaceutical re- effect in order to incur more broadly based manifestations search, development, testing and evaluation (RDTE) of and consequences. such agents engage the time and resources required to maximize desirable drug actions and effects based upon Neurotoxins individual and group geno- and phenotypes, and assess potentially adverse and/or unwanted side-effects. Of Of the aforementioned scenarios in which neuroweapons course, adverse effects could also be exploited for use on could be leveraged, the latter two are prime for the use of an enemy population. In light of this, drug design would organic neurotoxins. These agents are extracts or deriva- require resources necessary for evaluation and measure- tives of peptides found in mollusks (i.e., conotoxins), puff- ment of geno- and phenotypic characteristics that could be er fi sh and newts (i.e., tetrodotoxin) dinofl agellate algae optimized to selectively employ particular drugs within (i.e., saxitoxin), blue ringed octopus (i.e., maculotoxin), a population. By targeting these characteristics, it would and species of cobra (i.e., naja toxins). As depicted in Ta- be possible to mass deliver agents and still achieve some ble 3, all are potent paralytics, acting through mechanisms signifi cant measure of individualized effect(s). Current of ionic blockade (e.g., acetylcholine receptor agonism or efforts in “personalized medicine” may afford steps to- antagonism; or direct inhibition of sodium, calcium or po- ward realizing these possibilities (39-40). As well, certain tassium channels) in the peripheral nervous system and/ physiochemical obstacles to delivery have been overcome or at the neuromuscular junction (NMJ) (44), to induce by the use of nano-neurotechnologies that have allowed spastic or fl accid paralysis and cardio-respiratory failure. molecular conjugation or encapsulation of ligands, ligand Being peptides, the stability of these agents vary, but can precursors, and biosynthetic enzymes capable of cross- be enhanced through chemical alterations such as struc- ing the blood-brain and blood-cerebrospinal fl uid (CSF) tural cyclization, disulfi de bridge modifi cation, and sub- barriers, and thus permitting greater access to, and higher stitution of various residues (45), thereby increasing their bioavailability within the brain (1). utility.

Neuromicrobiological agents In all cases, paralysis occurs rapidly after introduction of a small dose of the agent, and the victim remains con- A number of microbiological agents directly target or indi- scious until overcome by shock and/or respiratory arrest. rectly affect the central nervous system (CNS), and these are As well, except for the naja toxins (for which there are a certainly employable as neuroweapons (1,10). Of particular number of species-specifi c antivenins, that each differ in note are 1) the viral encephalitides, such as the Alphavirus effectiveness), antidotes are not available, and rapid triage genus of the Togaviridae family that cause Venezuelan, east- for cardio-respiratory support is required to prevent mortal- ern and western equine encephalitis (41); 2) the anaerobic ity (although the effects of tetrodotoxin can be mitigated bacterium Clostridium botulinium, the seven strains of which with edrophonium (46). produce specifi c neurotoxins (42); and 3) the sporulating ba- cillus, B. anthracis that causes anthrax (42). (See Table 2)

T:64 Synesis: A Journal of Science, Technology, Ethics, and Policy 2011

Table 2. Neuromicrobiologic Agents

Category Agent Action Effects Viral encephalitides Togaviridae Alphavirus (e.g., Arthropod vectors (mosquitoes, Encephalitis: headache, nausea, Venezuelan, eastern, and ticks) vomiting, lethargy, seizures, western, and equine encephalitis Single-stranded RNA genomes coma, and possible paralysis or viruses Invade the brain via vascular focal signs of neuropathy. endothelial cells Bunyaviridae Orthobunyavirus Replicates in neurons of cerebral LCV: primarily children; (e.g., La Crosse virus (LCV) cortex <1% mortality; but most have and James Canyon virus (JCV) Causes neuronal necrosis persistent neurological sequelae variants of California encephalitis (e.g., recurrent seizures, virus) partial to full hemiparalysis; cognitive and neurobehavioral Flaviviridae fl avivirus (e.g., abnormalities) Powassan virus)

Microbial Amoebic parasite (e.g., Amoeba penetrate nasal mucosa Primary amoebic encephalitides trophozoites of Naegleria after insuffl ation or ingestion of meningoencephalitis (PAME) fowleri) infected water Progressive meningitis resulting Invades CNS via olfactory nerve. in encephalitis. Progressively necrotizes brain Death from respiratory failure tissue consequential to infl ammation and/or necrotization of the brainstem Mortality in 2 -4 weeks Protozoan parasite (e.g., Feline hosts Behavioral changes resembling Toxoplasma gondii ) those of dopamine reuptake Zoonotic transmission via human inhibitor type antidepressants contact or insuffl ation of fecal matter. and stimulants. Ingested; becomes active in Loss of impulse control; immunocompromised host agitation, confusion. Encodes the DA synthetic enzyme tyrosine hydroxylase Encephalitis Increases CNS DA levels Induces polyfocal neuroinfl ammation

T:65 Synesis: A Journal of Science, Technology, Ethics, and Policy 2011

Table 2. Neuromicrobiologic Agents, Continued

Category Agent Action Effects Bacterial toxigenics Bacillus anthracis (i.e., anthrax Bacterial spores are inhaled and Anthracic meningitis toxin) uptaken by macrophages Inhibition of protective immune Spores become active bacilli and responses rupture macrophages, releasing bacteria into bloodstream, where Cell lysis and destruction the anthrax toxin is released Bleeding and death Anthrax toxin enables bacteria to evade the immune system, proliferate, and ultimately cause polyfocal effects

Clostridium botulinum (i.e., Decreased acetylcholine release. Induces fl accid paralysis and Botulinum toxin) cardio-respiratory failure Decreased neuromuscular (also, C. butyricum transmission C. baratii C. argentinense) Several genera of cyanobacteria Stimulates nicotinic ACh receptors Permanent contraction of (e.g., anabaena, aphanizomenon muscles oscillatoria, microcystis, Mimics ACh causing irreversible planktothrix, raphidiopsis, stimulation of NMJ Loss of coordination, twitching, arthrospira,cylindrospermum, convulsions and rapid death by phormidium) (i.e., anatoxin-α) respiratory paralysis

Gambierdiscus toxicus (i.e., Lowers the threshold for activating Causes headache, muscle ciguatoxin) voltage-gated sodium channels aches, weakness, dizziness, itching. nightmares, and/or hallucinations. Causes parasthesias (e.g., sensation of burning or “pins- and-needles”, reversal of hot and cold temperature sensation, or unusual tastes) Very low mortality; recovery within 1 month.

Note: ACh: acetylcholine; DA: dopamine; NMJ: neuromuscular junction

T:66 Synesis: A Journal of Science, Technology, Ethics, and Policy 2011

Table 3. Organic Neurotoxic Agents

Origin Agent Mechanism Marine cone snails (genus Conus, σ-conotoxin Inhibits the inactivation of voltage-dependent sodium e.g., Conus geographus) channels; prolonged opening. κ-conotoxin Inhibits the inactivation of voltage-dependent sodium channels; prolonged opening. μ-conotoxin Inhibits voltage-dependent sodium channels in muscles.

ω-conotoxin Inhibits N-type voltage-dependent calcium channels. Symbiotic bacteria found in tetrodotoxin Prevents action potential fi ring in neurons rough-skinned newts, pufferfi sh Binds near the pore and blocks voltage-gated fast and procupinefi sh (e.g., sodium channels on presynaptic terminals Pseudoalteromonas tetraodonis, certain species of Pseudomonas and Vibrio) Symbiotic bacteria found in blue- Maculotoxin (a venomous form of Presynaptic sodium channel pore blockade ringed octopus (Hapalochlaena tetrodotoxin) Inhibition of action potential fi ring maculosa)

Shellfi sh/molluscs (e.g., Saxidomus Saxitoxin Selective pore blocker of neuronal voltage-gated giganteus contaminated by (e.g., saxitoxin, gonyautoxin, sodium channels algal blooms (“red tides”) e.g., neosaxitoxin) Water soluble and dispersible by aerosols Alexandriu catenella), and Inhalation causes death within minutes Pufferfi sh Cobra snake (Genus Naja) Naja toxins Block nicotinic ACh receptors at NMJ (e.g., α -cobratoxin) Also selective antagonist to α7-nicotinic ACh receptors in the brain

Krait snakes (Bungarus Bungarotoxins (e.g., Irreversible and competitive binding to NMJ nicotinic multicinctus) α-bungarotoxin) ACh receptors Also selective antagonist to α7-nicotinic ACh receptors in the brain Mamba snakes (Dendroapsis) Dendrotoxins (e.g., α-dendrotoxin, Block voltage-gated (A-type) potassium channels at σ-dendrotoxin) nodes of Ranvier in motor neurons Prolong duration action potentials Increase ACh release at NMJ

Australian taipan snakes Taipoxin Induced increasing blockade of presynaptic ACh (Oxyuranus scutellatus) release from motor neurons

Note: ACh: acetylcholine; NMJ: neuromuscular junction

T:67 Synesis: A Journal of Science, Technology, Ethics, and Policy 2011

Practical considerations, limitations and Still, neurotechnology will be evermore viable in translating preparations the nuances of social cognition and behaviors, and thus gain- ing a deeper understanding as to why certain principles The use of such neuroweapons, especially if apparent, is or violations are more or less likely to induce violence unlikely to result in lasting peace. Yet, the distribution of and/or strong opposition. Neurotechnology could— and a neurotropic drug or neuropathological agent through- likely will— play an increasingly larger role in exploring out a population could create a societal burden that sig- the relationships between culture and neuropsychological nifi cantly impacts the means, economic resources, and/or dynamics in and between populations. motivations to fi ght. But there is also the risk of a num- ber of “spillover effects”. First, given the environment(s) But, here we pose a caveat: Ignoring unresolved ambigui- in which most current warfare is conducted, it would be ties surrounding issues of the “brain~mind” and “reduc- nearly impossible to completely protect a civilian popula- tionist/antireductionist” debate (i.e., as connoted by the tion from the effects of neuroweapons. If the agent has a “neuro” prefi x) when employing scientifi c evidence as ra- known antidote or may be inoculated against, these might tionale for employing neurotechnologies as weapons may be relatively easy to counter—but this would depend upon lead to erroneous conclusions that may profoundly affect the integrity of the public health infrastructure of the town the intelligence and defense community – and the public. or country in which the agents are employed (and/or the An example is the current interest in using functional neu- capability of military forces to provide medical assistance roimaging (fMRI) for detecting deception (1). The valid- to those civilians that are affected). Second, if an antidote ity of this approach depends on the accuracy with which is not available, then there is risk of both serious collateral such technologies can detect psychological states relevant injury to the civilian population, and injury to one’s own to deception (such as anxiety versus something more ab- troops should they be exposed to the agent. Third, there stract, e.g., cognitive dissonance). How science portrays is risk to much broader populations if stocks of the agent the relationship between patterns that may be detected by were purloined from secure storage, or should a neuromi- neuroimaging, and what those patterns actually represent crobiological agent mutate while being employed (1). depends upon (neuroscientifi c) interpretation of the valid- ity of the technology (to actually do what it is intended), Evidently, these considerations prompt ethical and legal and, in light of this, the meaning of data and information, concerns that must be addressed – and resolved – through as constituting viable knowledge (24,49-50). the formulation of guidelines and policies. A complete discussion of the ethico-legal and social issues arising An illustration of how (mis-)conceptions of causal rela- from the use of neuroscience and technology in national tionships of the brain and cognition can constitute a ratio- security and defense agendas is beyond the scope of this nale for employing technologies or tactics is refl ected by essay (for a more detailed review of this topic, see the the following quote, from the May 4, 2009 issue of News- American Journal of Bioethics-Neuroscience, April-June week. In context, the speaker is asking the contractor who 2010, volume 1, number 2). Suffi ce it to state that any will replace him about a given approach to interrogation: and all use of neuroscience and neurotechnology in this regard mandates rigorous ethical analysis and discourse “…I asked [the contractor] if he’d ever interrogated that is inclusive of the research, academic, political and anyone, and he said ‘no, but that didn’t matter’, the public communities. contractor shot back, ‘Science is science. This is a be- havioral issue.’ He told me he’s ‘…a psychologist and… While the use of neurotechnology in national security, knows how the human mind works’” (51). intelligence and defense applications may be relatively new, the concept of using psychological science to de- This is relevant to the use of neurotechnology as it refl ects velop weaponry is not. In some ways, it may be as diffi cult a social tendency to concretize contingent neuroscientifi c to distinguish between neurotechnological and psychologi- understanding as “truth”. To be sure, knowledge in neu- cal warfare as it is to discriminate structure from function as roscience remains a work-in-progress, and thus, so does relates to brain and mind. “Changing minds and hearts” may any defi nition of what is real, true (i.e., based in fact), not be a task that is best addressed by neurotechnologies as valid, and possible as regards the applications and use(s) weapons. Instead, cultural sensitivity and effective commu- of neurotechnology. Still, neurotechnology can be used to nication might be a more desirable approach (1, 47-48). create weapons that may have an unprecedented capacity

T:68 Synesis: A Journal of Science, Technology, Ethics, and Policy 2011 to alter the cognitions, emotions, beliefs, and behaviors For this approach to work, surveillance (i.e., intelligence) of individuals, and groups – if not societies. Thus, the is necessary, and thus the development and use of many potential “power” of neurotechnology as weaponry lies of the aforementioned neurotechnological developments in the ability to assess, access and change aspects of a becomes increasingly important. Although international defi nable “self”. As with any weapons, they pose threats governance of neurotechnological RDTE may be diffi - to autonomy and free will, and can do so to an extent that cult, what can be governed and regulated are the ways psychological weapons alone could not. in which neuroscientifi c and neurotechnological RDTE efforts are conducted and employed by agencies of the It is foolhardy to think that the technological trend that US and its allies. In this regard, ethical questions need to compels the use of neurotechnology as weapons will be be prudently addressed and balanced with the interests of impeded merely by considerations of 1) the burdens and public (viz. - national) security and protection. risks that might arise as science advances ever deeper into the frontiers of the unknown; 2) the potential harms that Conclusion such advances could intentionally and/or unintentionally incur, and 3) the ethico-legal and social issues instantiated In an ideal world, science and technology would never be by both the positive and negative effects and implications employed for harmful ends; but we should not be naïve of these advances. This is because a strong driving (or and succumb to the dichotomy of ought versus is. Neu- “pushing”) force of both science and technology is the roscience can – and will – be engaged to effect outcomes human desire(s) for knowledge and control. At the same relevant to NSID operations by countries and non-state time, environmental events, market values, and socio- entities to achieve goals that are contrary to the interests political agendas create a “pulling force” for technologi- and public welfare of the United States and its allies. As cal progress, and can dictate direction(s) for its use. Both history has shown, a dismissive posture that fails to ac- former and latter issues are important to national security knowledge the reality of threat increases the probability and defense. In the fi rst case, the use of contingent knowl- of being susceptible to its harms. In an open society, it is edge could evoke unforeseen consequences that impact the responsibility of the government to protect the polis. public safety, and the power conferred by scientifi c and Hence, there is a duty to establish proactive, defensive technological capability could be used to leverage great knowledge of these scientifi c and technological capabili- power. In the second case, the intentional use of these ties and the vulnerabilities that they exploit, to recognize technologies by individual agents or groups in ways that how neuroscience and neurotechnology could be used to are hostile to the national interests of the US and its allies wage hostile acts, and to develop potential countermea- could incur profound public threats. sures to respond appropriately. But a meaningful stance of preparedness also mandates rigorous analyses and ad- Thus, a simple precautionary principle in which risk: ben- dress of the ethico-legal and social issues that such use efi t ratios determine the trajectory and pace of technologi- – and/or misuse – of neuroscience and neurotechnology cal advancement is not tenable on an international level, generate, and guidelines and policies must be formulated as there is the real possibility – if not probability – that to effectively direct and govern the scope and conduct of insurgent nations and/or groups could fund and covertly research and its applications in this area. Our group re- conduct RDTE of neuroweapons, beyond the auspices mains dedicated to this effort and approach. and infl uence of US (and/or UN) guidelines and policies. Instead, we argue for a process that entails some mea- Acknowledgments sure of precaution, together with signifi cant preparedness. Such preparedness requires knowledge of 1) what techno- The authors are grateful to Daniel Howlader and Rhian- logical accomplishments can be achieved (given the in- non Bower for their assistance on the fi nal preparation centives and resources afforded); 2) whether such work is of this manuscript, and acknowledge Sherry Loveless for being prepared and/or undertaken; 3) groups involved in the contribution of graphic artistry. such work; 4) overt and/or covert intention(s) and purpos- es; 5) what possible scenarios, effects and consequences Disclaimer could arise from various levels of technological progress, and 6) what measures can and should be taken to coun- The views expressed refl ect those of the authors and not ter threats imposed by such progress and its effects (52). necessarily those of their respective institutions.

T:69 Synesis: A Journal of Science, Technology, Ethics, and Policy 2011

Competing interests 15. Siegelmann H. Neural and super-turing computing. Minds and Machines. 2003;13(1):103-14. The author(s) declare that they have no competing interests. 16. Schemmel J, Hohmann S, Meier K, Schurmann F. A mixed-mode analog neural network using current- References steering synapses. Analog integrated circuits and sig- nal processing. 2004;38(2):233-44. 1. Committee on military and intelligence methodol- 17. Trautteur G, Tamburrini G. A note on discreteness ogy for emergent neurophysiological and cognitive/ and virtuality in analog computing. Theoretical com- neural research in the next two decades, National puter science. 2007;371(1-2):106-114. Research Council. Emerging cognitive neuroscience 18. Sato Y, Aihara K. A Bayesian model of sensory adap- and related technologies. Washington, DC: National tation. PLoS One. 2011;6(4):e19377. Academies Press; 2008. 19. Wiener N. Cybernetics: Or control and communica- 2. Merriam-Webster Dictionary. Weapon. 2008 [Inter- tion in the animal and the machine. Cambridge, Mas- net] [cited 2008 July 1]; Available from: http://www. sachusetts: Massachusetts Institute of Technology merriam-webster.com/dictionary/weapon. Press; 1948. 3. Romano JA, Lukey BJ, Salem H. Chemical war- 20. Clynes N, Kline M. Drugs, space and cybernetics: fare agents: chemistry, pharmacology, toxicology, evolution to cyborg. New York: Columbia University and therapeutics. 2nd ed. Boca Raton. Florida: CRC Press; 1961. Press; 2007. 21. Paddock AH. No more tactical information detach- 4. Abu-Qare AW, Abou-Donia MB. Sarin: health ef- ments: US military psychological operations in tran- fects, metabolism, and methods of analysis. Food and sition. In: Goldstein FL, Findley BF, editors. Psy- chemical toxicology. 2002;40(10): 1327–33. chological operations: principles and case studies. 5. McCoy AW. A question of torture: CIA interrogation, Montgomery, Alabama: Air University Press; 1996. from the cold war to the war on terror. Owl Books: p. 25-50. New York; 2006. 22. Lord C. The psychological dimension in national 6. Goldstein FL, Findley BF, editors. Psychological op- strategy. In: Goldstein FL, Findley BF, editors. Psy- erations: principles and case studies. Montgomery, chological operations: principles and case studies. Alabama: Air University Press; 1996. Montgomery, Alabama: Air University Press; 1996. 7. Moreno J. Mind wars. New York: Dana Press; 2006. p. 73-90. 8. Black J. The ethics of propaganda and the propagan- 23. Illes J, Racine E. Imaging or imagining? A neuroeth- da of ethics. In Wilkins L, Christians CG, editors. The ics challenge informed by genetics. American journal handbook of mass media ethics. New York: Rout- of bioethics. 2005;5(2):5-18. ledge; 2009. p. 130-48. 24. Uttal W. The new phrenology: the limits of localizing 9. Pringle RW, Ransom HH. Intelligence. In: Ency- cognitive processes in the brain. Cambridge, Mas- clopædia britannica [Internet]. 2009 [cited 2009 July sachusetts: Massachusetts Institute of Technology 26]. Available from: http://www.britannica.com/ Press; 2001. EBchecked/topic/289760/intelligence. 25. Ridley M. Nature via nature: genes, experience, and 10. Watson A. Neuromorphic engineering: Why what makes us human. London: HarperColins; 2003. can’t a computer be more like a brain? Science. 26. Giordano J. Integrative convergence in neuroscience: 1997;277(5334):1934-6. trajectories, problems and the need for a progressive 11. Konar A. Artifi cial intelligence and soft computing: neurobioethics. In: Vaseashta A, Braman E, Sussman, behavioral and computational modeling of the human P, editors. Technological innovation in sensing and brain. Boca Raton. Florida: CRC Press; 1999. detecting chemical, biological, radiological, nuclear 12. Von Neumann J. The computer and the brain. 2nd threats and ccological terrorism. New York: Springer; ed. New Haven, Connecticut: Yale University Press; 2011. 2000. 27. Hughes J. The struggle for a smarter world. Futures. 13. Giles J. Think like a bee. Nature.2001;410(6828):510-2. 2007;29(8):942-954. 14. Arbib M. The handbook of brain theory and neural 28. Hoag H. Neuroengineering: remote control. Nature. networks. 2nd ed. Cambridge, Massachusetts: Mas- 2003;423:796–8. sachusetts Institute of Technology Press; 2003.

T:70 Synesis: A Journal of Science, Technology, Ethics, and Policy 2011

29. Buguet A, Moroz DE, Radomski MW. Modafi nil— 44. Cannon SC. Ion-channel defects and aberrant excit- medical considerations for use in sustained opera- ability in myotonia and periodic paralysis. Trends in tions. Aviation, space, and environmental medicine. neurosciences. 1996;19(1):3-10. 2003;74(6):659-63. 45. Craik DJ, Adams DJ. Chemical modifi cation of cono- 30. Albucher RC, Liberzon I. Psychopharmacological toxins to improve stability and activity. ACS chemi- treatment in PTSD: a critical review. Journal of psy- cal biology. 2007;2(7):457-68. chiatric research. 2002;36(6):355-67. 46. Masaro EJ, editor. Handbook of neurotoxicology. To- 31. Davis LL, Suris A, Lambert MT, Heimberg C, Petty towa, New Jersey: Humana press; 2002. F. Post-traumatic stress disorder and serotonin: new 47. Freakley BC. Cultural awareness and combat power. directions for research and treatment. Journal of psy- Infantry magazine. 2005;94(March-April):1-2. chiatry & neuroscience. 1997;22(5):318-26. 48. McFate M. The military utility of understanding ad- 32. Gimpl G, Fahrenholz F. The oxytocin receptor sys- versary culture. Joint force quarterly. 2005;38:42-48. tem: structure, function, and regulation. Physiologi- 49. Illes J, editor. Neuroethics: defi ning the issues in the- cal reviews. 2001;81(2):629-83. ory, practice, and policy. Oxford: Oxford University 33. Murphy PN, Wareing M, Fisk JE. Users’ perceptions Press; 2006. of the risks and effects of taking ecstasy (MDMA): a 50. Farah MJ. Neuroethics: the practical and the philo- questionnaire study. Journal of psychopharmacology. sophical. Trends in cognitive sciences. 2005;9(1):34- 2006;20(3):447-55. 40. 34. Rubaj A, Zgodziński W, Sieklucka-Dziuba M. The epi- 51. Isikoff M. We could have done this the right way: leptogenic effect of seizures induced by hypoxia: the how Ali Soufan, an FBI agent, got Abu Zubaydah role of NMDA and AMPA/KA antagonists. Pharmacol- to talk without torture. Newsweek [Internet]. 2009. ogy biochemistry and behavior. 2003;74(2):303-11. April 25. Available from: http://www.newsweek. 35. Kautilya. Arthashastra. 3rd ed. Shamasastry R, trans- com/id/195089. lator. Mysore: Weslyan Mission Press; 1929. 52. Giordano J. Neurotechnology and national security 36. Thomsen K. A Foucauldian analysis of “a neuro- concerns: toward a precautionary process. Invited skeptic’s guide to neuroethics and national security.” lecture; 2009 September 16; conducted at the Center AJOB neuroscience. 2010;1(2):29-30. for National Preparedness, University of Pittsburgh; 37. Rippon G, Senior C. Neuroscience has no place in na- Pittsburgh, Pennsylvania. tional security. AJOB neuroscience. 2010;1(2):37-38. 38. Foucault M. Security, territory, population: lectures at the Collège de France 1977-1978. New York: Pal- grave Macmillan; 2007. 39. Calabrese EJ, Baldwin LA. Hormesis: U-shaped dose responses and their centrality in toxicology. Trends in pharmacological sciences. 2001;22(6):285-91. 40. Nebert DW, Jorge-Nebert L, Vesell ES. Pharmacog- enomics and “individualized drug therapy”: high ex- pectations and disappointing achievements. American journal of pharmacogenomics. 2003;3(6):361-70. 41. Smith JF, Davis K, Hart MK, Ludwig GV, McClain DJ, Parker MD, Pratt WD. Viral encephalitides. In: Sidell FR, Takafuji ET, Franz DM. editors. Medical aspects of chemical and biological warfare. Washing- ton, DC: Borden Institute; 1997. p. 561-89. 42. Nagase H. Metalloproteases. Current protocols in protein science. 2001;21(4):1-13. 43. McGovern TW, Christopher GW. Cutaneous manifes- tations of biological warfare and related threat agents. Archives of dermatology. 1999;135:311-22.

T:71