Roundtables
ShareThe military application of neuroscience research
Defense officials have dreamt up a range of uses for neuroscience research aimed at monitoring brain function and affecting brain performance in humans and animals alike. Yet, the impact of this research is just beginning to become apparent. Georgia Tech researchers Jonathan Y. Huang and Margaret E. Kosal investigated the many implications of this research in "The Security Impact of the Neurosciences." Below, Kosal and three colleagues delve further into the implications of these trends in neuroscience research and discuss how to stem the technology's misuse.
I'd like to change the topic for a moment. What strikes me about the reaction to the National Research Council report ("Emerging Cognitive Neuroscience and Related Technologies") in the past few days is the attention that is being paid to the intersection of neuroscience, conflict, and culture. Both traditional philosophers and pioneers in neuroscience have long suspected that there may be differences in the way people from different cultures both process information and the way they understand what it is to process information. The former point is a bit more straightforward but could be attributed to behavioral psychology (a tradition that runs from St. Augustine to B. F. Skinner), or evolutionary biology (including E. O. Wilson's sociobiology). There is no reason to rule out the possibility that powerful new techniques for reading many individual genomes could develop the field of "comparative cultural genomics" so that a physical substrate of these differences could be identified. Whatever degree of accuracy this data could provide would be of enormous interest to both national security officials and diplomats.
The second point calls to mind an M. C. Escher work. Perhaps the way different cultural groups typically understand what they are doing when they are, fill in the blank--experiencing, learning, reflecting, deliberating, praying, or otherwise engaged in some form of reverie--itself has an influence on the way information is processed and stored. What I have called comparative cultural genomics could tell us whether that is the case or not. But if it did, the task would seem to be incomparably more complicated if there were serious doubts about whether our science-based system for understanding neural activity is commensurable with other modes of understanding.
In other words, we might end up right back where we started. The most efficient option might be to "go native" rather than attempt some translation via neuroscience. And the only way to confirm our success would be to get along with the natives when we expect to, and not when we don't expect to. This point was brilliantly made by W. V. O. Quine in his classic, "Word and Object." In the end, our ability to know how others know is a theory, however sophisticated its expression, and what stands behind the theory remains a question of philosophical wonder.
I would like to reflect on the approaches taken by my colleagues, Jonathan Moreno and Margaret Kosal, in the beginning of this discussion. Specifically, they have addressed what they contend to be two important core "rate-limiters" in the future progress of neuroscience research that could have military application: funding and ethics. While I do not disagree with anything that Jonathan and Margaret say in principle, my view is that these two core issues make it clear that this discussion must become global. If not, we will miss the opportunity to address the fundamental issue: what are the military applications of neurosciences? The research conditions in U.S. labs are well known, and much if not most of the important militarily relevant neuroscience research will take place overseas. If this discussion becomes global and we engage with non-U.S. researchers, then we can really begin to address this question.
While Jonathan, Margaret, and I are all in violent agreement about the basics, we need to discover the subtle differences in our opinions to help us learn. To do this, I will purposely overstate in tone (not in facts) two key issues: First, I stand by my initial belief that the pace of discovery in the neurosciences is unlikely to accelerate faster than it is today and will not happen preferentially in the West. Second, we should not assume that the military applications of this research can be influenced by incremental additional financial support.
The pace of discovery in neurosciences in the West is widely believed to be driven by stable and moderately well-funded Centers of Excellence. The research with applicability to military use is well known and well publicized. Those of us with access to information about classified programs are underwhelmed by the lack of significant novel research. In fact, as part of the National Academy of Sciences commission I head, the 16 members of the committee, 6 staff, and 12 peer reviewers humbly received briefings and work product from 25 military and intelligence agencies, national laboratories, universities, nongovernmental organizations, and other private institutions doing neurosciences research directly or indirectly involved in work of potential military value, and reviewed hundreds of peer-reviewed publications. The committee decided that it had no scientific or substantive reason to write a classified annex. (The peer review of the commission's final report is complete and the report will be released as "Emerging Cognitive Neuroscience and Related Technologies.")
From the information that is accessible, we know the ongoing work is relatively well supported, as Margaret described. Should we expect any significant increase in a few years for "basic" research? The bigger problem is that the military and intelligence communities cannot understand the implications of any such research. As shocking as it may be, fewer than a dozen persons in intelligence and military constituencies understand the neurosciences involved, and they are happy to say so--hence, the plethora of outside reviews on the subject in the past two years. And government scientists are not well placed to accelerate application development from the basic neurosciences. Margaret has it just right: Only some form of increased exchange and communication between academia and the government in this area will work.
Additional funding alone won't help, in any event, for a separate reason. The goal of the basic neurosciences funded today is for the most part hypothesis testing, not hypothesis generating. Grants and awards are also appropriately constrained by the Health Insurance Portability and Accountability Act (HIPAA) and ethics. As Jonathan wisely pointed out, the "first experiments" in the neuroscience arenas most talked about--psychopharmacology and aides to interrogation--have to wait for approvals and reviews. The drugs and compounds being considered today are older, and the ones we might expect to be useful to modify emotional state will not be tested on humans in any context of reasonable funding or ethics. It is not in the realm of possibilities that next-generation drugs with safe and efficacious properties for military use will be developed absent a huge increase in funding. I am happy to go on record as stating that for both scientific and ethical reasons I oppose the development of such drugs.
Funding priorities require a solid roadmap that includes research that is certain to meet prescribed goals as contained in the rules of grant applications. From the landscape we have seen, it is unlikely that a disruptive technology could escape from a basic science laboratory. (Clayton Christensen, thank you for teaching us that disruptors by definition are today's technology turned inward). With careful planning and human-use approvals, it is just as unlikely that a new discovery with unintended consequences will pop-out to become a new military application, unless it is looked for systematically--almost (but not quite) an oxymoron. Screening for the unintended applications of a drug is not finding a surprise, it is reading carefully the data from good epidemiology.
It would be productive to turn this discussion. Basic science funding is likely to stay stable and little serious research will delve into the psychopharmacology of interrogation. Thus, a vector that we may wish to explore is the new data from several global laboratories that indicates culture matters in the decision to use any military application of asymmetric force. Battlefield commanders of all nations hold sacrosanct the right to determine the applications that may cause harm to those outside the bandwidth of an lethal dose 99+ weapon and generally don't intend to develop for use materials that could cause collateral harm to civilians and non-combatants. If governments or scientists were to try to develop a system to pre-screen neuroscientific cognitive manipulators, which would be HIPAA approved and tested, and robust in its core science, success would be as likely as it was with mines and cluster-bombs--meaning not likely. And if we did have such success, our enemies of the future would not care.
Thanks to my colleagues for furthering this conversation. I look forward to the results of the National Academy of Sciences study that Christopher Green is chairing.
The intersection between the cognitive sciences (of which neuroscience may be seen as a subset) and national security offers many puzzles--scientific, ethical, policy, and practical. Jonathan Moreno's comments highlight the critical need for more research into the underlying physiological mechanisms of proposed neuropharmocological chemicals and their pharmacokinetics (how and how fast they are distributed, absorbed, processed, and excreted by the body). It is a challenge to understand the effect of a pharmacological chemical across a wide population of individuals, which would complicate using such an agent to deal with a notional hostage or insurgent situations. An aerobically fit 25-year-old male is likely to be affected very differently than a 75-year-old man, a 32-year-old pregnant woman, or an 8-year-old girl. The challenge and cost increases if each neurochemical has to be tested individually, and can be further confounded by the condition of subjects, such as increases in adrenaline or exhaustion, to which Moreno alluded. These challenges highlight the need for more basic research into the chemistry and biology of the brain and for the development of predictive models to understand and predict the underlying phenomena. To realize such research, robust funding for research in the basic sciences is critical.
The Defense Department has invested substantially in the scientific research community, including basic research support to academia. According to data publically available through the 2008 Presidential Budget Request, the Defense Advanced Research Projects Agency (DARPA) invested at least $372 million in the cognitive sciences in 2007, including areas such as cognitive computing and bio-revolution. Other basic research investments in cognitive sciences across Defense included: $13.3 million for the Army, $13.9 million for the Air Force, and $10.4 million for the Navy in 2007.
Outside of Defense, the two significant funders of basic research in the cognitive sciences are the National Science Foundation (NSF) and the National Institutes of Health (NIH). The NSF, with a nearly $6 billion budget in 2007, maintains active grant programs in perception, action, and cognition; cognitive neuroscience; neural systems; and collaborative research in computational neuroscience. The NIH reported $1.8 billion in appropriation in 2006 for its two agencies most relevant to brain research: the National Institute of Bio-Imaging and Bioengineering and National Institute of Neurological Disorder and Stroke. It's less clear how non-defense federal agencies appropriate basic research funding in the cognitive sciences.
Methods for inter- and intra-agency research coordination, as well robust public oversight, are needed across cognitive sciences research investment. Public oversight is particularly notable--and challenging--as it applies to potential applications of ethical concern, whether they are security related or not. When talking about the potential implications of emerging sciences such as the cognitive and neurosciences, partisan or political overtones can sometimes infect a discussion or policy. To counter this tendency, there is a critical need to ground possible scenarios in technical viability. Christopher delineates between the "achievable" and "unachievable" scientific goals, and this is one area that greatly concerns me.
Compared to biotechnology and even nanotechnology, civilian and military applications of cognitive sciences are nascent. Underlying some discussions of the applications of nanotechnology have been notional scenarios that fall into the "unachievable" category, e.g., concerns about self-replicating abiotic molecular assemblers and calls for international treaties limiting their production and use. These scenarios speak to the critical need for scientists and engineers to be involved in policy formation and public dialogue. Although stating that to readers of the Bulletin is speaking to the metaphorical choir.
Jonathan Huang and Margaret Kosal have prepared a helpful typology of neuroscience applications to national security functions. Although there is much to be said concerning each of the three categories they identify, I will confine this comment to neuropharmacology.
Calmatives agents are in theory an attractive approach not only for national security purposes but also for domestic policing. However, the time lag between the release of a substance and its effect on targeted individuals makes currently available opiates such as fentanyl poor candidates for such uses. It is not well understood, for instance, why the group holding the Moscow Theater hostages did not react when it became apparent that something was going on in the confines of the building, but perhaps they were so overcome with exhaustion that they did not register events quickly enough. Whatever the explanation, there is no assurance that future hostage-takers bent on suicide could be managed in this way. Governments will need to develop a faster acting agent before using calmatives in this way is an attractive option. Also, the open-air release of an agent is unlikely to be effective due to the dispersal of the aerosol, limiting the situations in which a calmative can be an effective non-lethal weapon.
Nor should one assume that even effective calmatives would truly stem violence. Reaction to their use might well stimulate aggressive reprisals by adversaries, so that ultimately in the course of a conflict there would be no reduction in violence. We should resist the illusion that there can be any technological fix to the hostilities inherent in a zone of combat.
The enhancement possibilities mentioned by Huang and Kosal should prompt us to recall that there must be a first user, and that that first use should be considered a clinical experiment. Governments have historically introduced performance-enhancing drugs to soldiers based on little evidence of their efficacy as compared to the tradeoffs, e.g., artificially extended wakefulness versus impaired judgment and reflexes (cocaine, caffeine, amphetamines). The effects of substances like modafinil, a stimulant, outside the laboratory need to be carefully examined in environments that simulate potential combat situations. The larger political and social question is how much enhancement future warfighters can legitimately be expected to accept as part of their preparation for service, especially since it will be some time before the long-term effects of neuropharmacology can be understood.
Not mentioned by Huang and Kosal are the possibilities that neuropharmacology may present for enhancing the accuracy and efficiency of interrogation procedures. In at least a few cases, the United States has used techniques that have conventionally been considered torture to obtain information from persons detained as suspected terrorists. A few days before these comments were written, documents surfaced in a U.S. Senate hearing on the treatment of detainees that tie techniques used against American POW’s during the Korean War by Chinese interrogators to an interrogation manual used by the U.S. military at Guantanamo Bay. The U.S. manual was based on a 1957 analysis of alleged "brainwashing" by the Chinese that resulted in false confessions of crimes by U.S. soldiers.
Public outrage about such revelations would seem to justify interest in chemical approaches to interrogation as an alternative to overtly violent techniques, for example, the use of a version of oxytocin to stimulate a trusting response in an individual. Research suggests that such an effect could be evoked from artificially stimulated neurochemical production or from drugs. It is a good bet that there will be attempts to develop such substances. In an odd way we may be about to circle back to the 1950s notion that LSD or other hallucinogens could be "truth serums." We need to tread carefully here. Although less overtly violent than techniques like water boarding, directed chemical changes in human consciousness under duress are subtly powerful invasions of the personality. As we well know from the experience with atomic weapons, once these genies are released under the acute duress and justification of conflict they are hard to put back.
Jonathan Huang and Margaret Kosal have done an excellent job in carefully framing this discussion. They courageously include mention of how "existing international agreements are inadequate to address the security implications of neuroscience research." The very notion of international agreements leads to an implicit belief that traditional arms control approaches could be relevant to this domain, when in fact, they are irrelevant. The pace of development of the technical areas--neuropharmacology, neuroimaging, and brain-machine interactions--that Huang and Kosal chose to address will outpace the hysteresis of the ponderous and arcane processes of traditional security control and disarmament.
The contents of a soon to be released National Academies study, "Military and Intelligence Methodology for Emergent Neurophysiological and Cognitive/Neural Science Research in the Next Two Decades," undertaken by a 19-member panel that I chair, will complement the scope that Huang and Kosal propose. The study does not address arms control issues directly, yet implicit is ample evidence that in the next 20 years, the pace of development of neuroscience technologies related to the military and intelligence communities will swamp traditional arms control measures. Even in the domain of biomedical ethics, the one area that is most relevant to arms control, the rights of individuals are bound in cultural variability and are likely to be out of reach for agreements.
In discussing these issues, it is possible for this roundtable's dialogue to stray toward a "liberal versus conservative" view of science, the brain, and the world. That would be unfortunate, and would likely result in a one-sided discussion, as people who work daily in the laboratories and hospitals doing emergent neuroscience clinical research would surely leave the discussion. The ethics issues at stake are real and worrisome; the chances of agreement on solutions are zero. Yet, the chances of agreement on the challenges are high. The pace of global research in the three areas under discussion (and additional areas discussed in detail in the Academies' report) is even faster than is reported today in the mainstream press, which sinfully exaggerates what is really going on. Another point of agreement will be the really worrisome work being done outside the public eye.
We need a fresh sort of discussion to address these challenges. Four noncontroversial matters make the case for this new approach:
- Several of the cognitive science applications that have the most enormous ethical implication (as described by Huang and Kosal) contain advances that are being made even faster than publicly thought. These advances do not include lie detection technology, whose potential to invade the privacy of individuals is an unrealistic scientific possibility (I'm waiting for a theory of mind to be developed first!). They do include an approach to near-real-time, multimodal cognitive measurements to "watch people think" while under stress (an achievable goal scientifically) not under duress (an unachievable goal scientifically).
- Big Pharma is global. Drug discovery research is both ponderous (not as much as arms control, however) and increasingly beyond the control of governments and the public. The development of cognitive enhancers and anti-aging aides during the next two decades (the time needed for drug discovery to become successful) will be identical to what Huang and Kosal call out as ethically worrisome. But it will be beyond opprobrium. Drugs will be developed and marketed, and not necessarily under the auspices of traditional Western controls and good laboratory practices.
- In the past, clinical and experimental psychology have been relatively easy to follow and understand for educated laypersons and life scientists, at least compared with biochemistry and biophysics, genomics and proteomics, and drug design and development. In the future, psychology experiments will increasingly be done in neuroimaging laboratories, which will be dependent on new physics, imaging scan sequence developments, and more complicated experimental designs. The work will be paid for by pharmaceutical companies or national militaries, bringing advanced signal processing and advanced mathematics into the experimental psychology laboratory. Not all of the "good" research will be done exclusively in the West--at least not before 2015. The results of this work will make sick people well and soldiers safer, but the technologies will not exclusively follow Western views on ethical questions, such as human stem-cell research, research on willing prisoners, and work on human-animal chimeras. The shrinking differences between the science of neural discovery aimed at helping sick people and the work being done to enhance soldier performance will likely dominate policy discussions at the end of the day. Notably, this will be seen in human and operator performance enhancement and neuropharmacology, as addressed by Huang and Kosal.
- Individual scientists in the areas of cultural anthropology and ethnography will soon be trained in the physics of neuroimaging and neuropharmacology. Thus, learning about why people misbehave will be subjugated to learning new methods of unethical manipulation of innocent peoples' behaviors. This roundtable may give a new and frightening meaning to the already scary 20-year-old dual-use construct.