Comments
I recently read (most of[1]) the 2014 book Barriers to Bioweapons: The Challenges of Expertise and Organization for Weapons Development by Sonia Ben Ouagrham-Gormley. (See also the author’s 2012 paper on the topic and this book review)
I found the book very insightful and recommend it to others interested in biosecurity, especially those interested in biological weapons nonproliferation and the debate around life science dual-use research of concern. The book is well-written, highly accessible, and fairly short (~170 pages).
Barriers to Bioweapons is a valuable corrective to the common but mistaken belief that (given current technology) it is trivially easy and cheap to develop biological weapons. Its key argument—that tacit knowledge presents the main obstacle to biological weapons development—has important policy implications for those seeking to reduce biological risks. Moreover, the book offers a fascinating perspective on the scientific process and innovation beyond the life sciences.
While I disagree with some of the book’s conclusions, the book is nevertheless well worth reading. My main disagreements concern the author’s opposition to publication restrictions and her optimism that future biotech progress won’t drastically erode the ‘barriers to bioweapons’.
In this post, I share some of my copied-and-pasted notes from the book that were of particular interest to me (excluding chapters 4-6). These notes cannot do justice to the book's extended discussion. Still, I hope some people may find my notes valuable and perhaps be inspired to read the book.
To let you decide whether to keep on reading, here is a summary of the book’s core argument (taken from here):
In both the popular imagination and among lawmakers and national security experts, there exists the belief that with sufficient motivation and material resources, states or terrorist groups can produce bioweapons easily, cheaply, and successfully (...) Sonia Ben Ouagrham-Gormley challenges this perception by showing that bioweapons development is a difficult, protracted, and expensive endeavor, rarely achieving the expected results whatever the magnitude of investment (...)
Bioweapons development relies on living organisms that are sensitive to their environment and handling conditions, and therefore behave unpredictably. These features place a greater premium on specialized knowledge (...) lack of access to such intellectual capital constitutes the greatest barrier to the making of bioweapons (...) The specific organizational, managerial, social, political, and economic conditions necessary for success are difficult to achieve, particularly in covert programs where the need to prevent detection imposes managerial and organizational conditions that conflict with knowledge production.
Throughout this post, my highlights are in bold.
Interviews with former bioweaponeers conducted in the United States, Russia, and Kazakhstan between 2008 and 2012 told a story about past bioweapons efforts that requires us to reconsider current threat assessments. Put simply, their testimonies show how difficult, protracted, and expensive bioweapons efforts have been; outcomes rarely achieved what the magnitude of investment might have suggested. This is good news for nonproliferation. The bad news, however, lies in our ignorance of key determinants of bioweapons development that create new opportunities for proliferation
Would access to published documents suffice to allow replication of past work? If so, does this mean that the bioterrorism threat automatically increases with the progress of science? In this book I argue that the answer to these two questions is negative.
This is contrary to the belief, shared by many analysts and policymakers, that bioweapons development requires only the procurement of three easily accessible resources: biomaterials, scientific data, and equipment. Therefore, the question of what skills and what conditions would allow replication is not considered.
Yet the analysis of past state and terrorist bioweapons programs shows that producing a working bioweapon is not a simple process of material accumulation. The challenge in developing biological weapons lies not in the acquisition but in the use of the material and technologies required for their development. Put differently, in the bioweapons field, expertise and knowledge—and the conditions under which scientific work occurs—are significantly greater barriers to weapons development than are procurement of biomaterials, scientific documents, and equipment.
Contrary to popular belief, however, this specialized bioweapons knowledge is not easily acquired. Therefore, current threat assessments that focus exclusively on the formative stage of a bioweapons program and measure a program’s progress by tracking material and technology procurement are bound to overestimate the threat.
measuring the pace and success rate of a program requires a detailed understanding of what factors shape knowledge acquisition and use within that program.
Finally, because the variables that truly affect the success of a bioweapons program are not currently addressed, the door remains open to proliferation. This reality dictates major changes in nonproliferation and counterproliferation approaches to address actual bioweapons threats. Current policies focus almost exclusively on preventing access to the troika of resources deemed essential for bioweapons development: material, scientific information, and technologies. By also targeting knowledge and the factors that affect its use, these policies could more effectively inhibit the growth of a weapons program and possibly bring about its collapse.
Three misconceptions are at the heart of the current faulty assessment of the bioweapons threat. The first finds its roots in the use of the nuclear model as a starting point to assess bioweapons development. Put simply, because biological weapons do not face the same stiff material barrier as do nuclear weapons, they are deemed easy and cheap to produce. The second lies in the assumption that any biology-related knowledge is applicable to bioweapons development and that bioweapons expertise is easily acquired and used. The third assumes that new technologies will erase the technical barriers to bioweapons development, allowing even untrained individuals to achieve successful results.
[nuclear and biological weapons] use materials of a decidedly different nature, which create barriers to entry at different points of their development process (...)
In the nuclear field, a key barrier to entry is located at the front end of the development process, at the stage of material acquisition. Achieving nuclear weapons is indeed conditioned by the ability to produce fissile material, which requires large facilities and specialized equipment.
When applied to bioweapons, however, the front-end/material- based nuclear model produces a distorted and even apocalyptic picture of the threat. Most analysts and policymakers stress that pathogens—viruses, bacteria, and toxins—can be isolated from nature or obtained commercially because they also have legitimate commercial or pharmaceutical use. They point out that equipment is essentially dual use and can therefore be readily purchased, while scientific publications provide ample descriptions of experiments and techniques that many believe can be easily replicated. (...)
Unlike nuclear weapons, which rely on materials with physically predictable properties, bioweapons are based on living organisms or by-products of living organisms, which evolve, are prone to developing new properties, and are sensitive to environmental and handling uncertainties. Their behavior, therefore, is unpredictable throughout all stages of development and use as a weapon, which imposes an extended trial-and-error process to acquire the skills necessary to solve problems that inevitably arise. (...)
unlike nuclear weapons, which can use only two sources of material—highly enriched uranium and plutonium-239—bioweapons development can exploit a large number of agents that vary in type (bacteria, viruses, toxins), properties (virulence, infectivity, transmissibility), and ease of culturing. Each agent also includes several strains with varying properties, further complicating bioweapons development.
If bioweapons developments were so simple, more states and terrorist groups should have achieved satisfactory results. But historical evidence shows otherwise. (...)
none of the past bioweapons programs have been completely successful. The Soviet Union, which had the largest and longest-running program, did not reach the level of accomplishment that its sixty-year lifespan and estimated investment of $35 billion might suggest. Soviet scientists successfully weaponized several classical agents, loading them into a variety of bombs, but according to recent evidence, their work on engineered pathogens—the program’s main focus during its last two decades—did not extend beyond the exploratory phase.
The American program, arguably the second largest program after the Soviet Union’s, cost about $700 million over twenty-seven years but resulted in only a small arsenal of bombs filled with half a dozen agents, and no ballistic or cruise missiles to deliver them.
Other states and terrorist programs performed even more dismally. Iraq invested twenty years and over $80 million during the last five years of the program alone to produce ineffective bombs that would have destroyed most of the liquid agents they contained. South Africa devoted twelve years and over $30 million to its program while producing only poisonous substances for assassination purposes. Finally, the Japanese terrorist group Aum Shinrikyo spent six years and about $10 million trying to produce anthrax- and botulinum-based weapons but failed at every stage of these bioweapons’ life cycles.
working with live organisms is not easy. (...) this unpredictability places greater emphasis on possessing the unique skills necessary to handle highly capricious biological agents and maintain their desirable properties throughout the development process. (...) Due to the fragility of living microorganisms, possessing the skills to handle and manipulate them throughout the development process is a greater barrier to entry into the bioweapons field than is material procurement.
these challenges are particularly acute as the agent moves down the development process toward weaponization. (...)
One such stage is scale-up. Biomaterials do not scale up easily. Yet the passage from a laboratory sample to larger quantities, whether a few gallons used in a terrorist program or industrial quantities used in a state program, stands as a critical stage of bioweapons development. Because scale-up is not a linear process, increases in quantity must occur gradually. Each increase, however, entails new challenges that impose changes to the production protocols. (...)
production and scale-up often subjected bioagents to contamination, which caused multiple delays and failures both in the U.S. and the Soviet bioweapons programs. Pharmaceutical and biotech companies also routinely endure such failures due to the complexity and sensitivity of biological organisms. (...)
Successful scale-up also requires the intervention of interdisciplinary teams with a variety of skills, whose work must be carefully integrated and coordinated. Integration and coordination are particularly important in large programs, which may involve thousands of individuals and hundreds of facilities (as in the Soviet program), but also in smaller ones, such as the Iraqi program, which involved about one hundred people.
the expertise acquired while working with one agent does not necessarily transfer to another microorganism. This lack of knowledge transferability limits the ability of a state or group to transition easily to new types of bioweapons should their work with one agent fail to produce tangible results. (...) transitioning to a new organism entails a lengthy period of knowledge acquisition, which would necessarily delay progress.
The second tenet of the current view of bioweapons proliferation is that science-based knowledge and technology are universal, independent of context, impersonal, public, and cumulative. As a result, knowledge spreads easily, and written documents, such as scientific publications, weapons designs, or scientific protocols, constitute complete representations of a technological artifact, thus allowing the replication of past work, even by untrained individuals.
however, knowledge is far from free flowing. Studies of knowledge transfer in various technological environments, including bioweapons technology, show that access to scientific documents does not guarantee their successful use, even by experts. This is so because scientific documents include only a small fraction of the knowledge produced within a program or scientific experiment.
technical knowledge results from a process of experimentation that produces both explicit and tacit knowledge. Whereas explicit knowledge can be codified and encapsulated in different physical formats that are easy to transfer—protocols, formulas, or designs—tacit knowledge is constituted of know-how: unarticulated skills or practices that cannot be reduced to a written form and are often personal, local, and context specific. Their transfer requires direct and prolonged interaction among people, and their use in a new location requires adaptation to the new site. (...) unlike explicit knowledge, which has a long shelf life, tacit knowledge decays over time if not sustained through practice.
complex projects such as bioweapons development involve teams of scientists and technicians, representing different disciplines, who, through their interactions and cooperative work, produce a different form of knowledge—known as communal knowledge—which is shared by all team members but possessed entirely by none. As a result, even individual experts possess only a limited knowledge of a whole weapon and its development.
Replicating past work using scientific documents alone, therefore, cannot be achieved without access to the corresponding tacit skills and the related communal knowledge. (...)
Tacit knowledge aside, written or published scientific documents are also often incomplete because they fail to emphasize essential aspects of scientific success, such as the contingencies associated with key stages of an experiment, the characteristics of the equipment—some scientists have their equipment custom-made to increase the rate of success for certain manipulations—or the laboratory routines that constitute essential parts of the scientific discipline and success. Experimental work also involves making changes on the fly that are not necessarily recorded, either for lack of time or because the scientists do not recognize the importance of these changes for experimental success. Without such details, reproduction is made even more complicated. (...)
Thus, scientific innovations and their corresponding documents require extensive interpretation and judgment from users other than their authors, which in turn necessitates the possession of prior base knowledge and ideally the ability to work in close cooperation with the documents’ authors. In addition, applying these scientific findings to bioweapons developments requires the acquisition of bioweapons expertise, which can take years. These conditions constitute major obstacles for replication of past work, particularly for untrained individuals operating under covert conditions.
concerns about the risk associated with publication of dual-use research hinge on two other erroneous assumptions. (...)
First is the assumption that expertise acquired in a civilian laboratory can easily be applied to bioweapons work. Here again history belies this belief. (...)
The second assumption is that innovations achieved in the laboratory can be easily fashioned into a harmful agent or a bioweapon. In fact, past bioweapons work, as well as current pharmaceutical efforts, show that transforming a scientific concept developed in the laboratory into a product that has a specific applied purpose and that functions reliably and effectively can take several decades and require the intervention of a wide array of expertise (...) bioweaponeers also face the challenge of developing a delivery mechanism that will protect living organisms and toxins from degradation due to environmental conditions (...)
Therefore, laboratory successes are not equivalent to successful application to a specific purpose. Specialized skills acquired through hands-on involvement in production and weaponization work are needed.
Because they automate processes that previously required the manual intervention of skilled personnel, new technologies are believed to facilitate replication of past work and allow their application to bioweapons development. What’s more, economic globalization and the rapidly decreasing cost of such technologies are deemed to be key factors in accelerating their diffusion, making it almost impossible to rein in bioweapons proliferation. (...)
The narrative behind the biotechnology revolution rests on two important premises:
(1) new technologies and equipment are black boxes, with an input and an output, which can be used by any user, irrespective of their technical skills; and
(2) technology developments result in a gradual deskilling of technological and scientific work. Therefore, machines and their accompanying instructions become the embodiment of human knowledge, captured and codified for easier use by less experienced individuals.
If knowledge was completely embedded in machines, all users would achieve equal results whatever their level of expertise. Empirical research shows that this is not the case. (...)
Instead, technologies, very much like written documents, constitute imperfect representations of their designers’ knowledge, and although they may simplify some tasks, they require from their users extensive experimentation, interpretation, and adaptation to a new location to achieve successful results. This is due to the fact that new technologies rarely automate all aspects of a task, requiring scientists to perform some tasks manually. (...) Furthermore, machines, however sophisticated, are also prone to errors, requiring their users to possess the skills to identify and correct problems.
My analytical framework defines the sustenance phase of a program as the key stage of bioweapons development, in which knowledge acquisition is the key variable. I also offer two sets of factors that affect the use of knowledge—those from within a program (endogenous variables) and those from the outside (exogenous variables)—the combination of which results in different speeds and outcomes. (...)
Most important to a program’s timeline and outcome is the way in which states or terrorist groups integrate these two sets of variables: social, organizational, and managerial factors that influence knowledge acquisition and scientific work from within; and exogenous variables that influence achievements from the outside. (...)
endogenous and exogenous characteristics of a bioweapons program can facilitate or hinder its development. Although access to material resources is important, it is the combination of organizational, managerial, political, and economic circumstances characterizing a program that ultimately affects its ability to produce and use knowledge, and thus affect the pace and ultimate program output.
technology results from the interaction among individuals who combine their respective expertise to produce a working technological artifact. A program’s success will therefore depend on these individuals’ ability to cooperate, exchange information, learn from one another, and institutionalize knowledge. (...)
Because knowledge transfer depends on the quality and frequency of individual interactions, the manner in which a program organizes and manages its resident-expert knowledge will also influence scientific outcomes.
Consequently, a bioweapons program’s structural and work organization, its management style, and the social context within which knowledge is created constitute the endogenous variables that figure into a program’s success or failure.
weapons programs do not happen in a vacuum; they frequently depend on their external context. Several such exogenous variables influence the conditions in which scientific work occurs.
Foreign technical assistance is one of these variables. (...) Two other exogenous factors—the priority political or group leaders devote to a program, and a program’s economic circumstances—also affect a program’s outcome, not only because they have an impact on programmatic and funding decisions but also because they influence the continuity and stability of scientific work—two essential conditions for the accumulation of knowledge. These factors may also generate scientific behaviors that can produce bad science. (...) Finally, in some cases, the location of a program can promote or constrain its development and the successful use of technologies by affecting the properties of the material used in laboratory work.
covert [bioweapons] programs face greater limitations on information exchange, organization, management, and their ability to deal with exogenous factors.
specialized knowledge is not easy to acquire, use, or transfer because much of it is tacit, local, and collective in nature. In addition, scientific knowledge pertaining to a technological artifact is rarely stored in a central location. It is more often held in various interdependent reservoirs, each containing a portion of the relevant knowledge. Moreover, knowledge decays over time if it is not used or transferred to the next generation. These factors raise questions about whether an untrained terrorist could in fact produce a biological or some other weapon of mass destruction (WMD) using easily accessible scientific data and material. (...)
The political science and policy literature is filled with assertions that knowledge and science-based technology are universal, independent of context, impersonal, public, and cumulative. This view suggests that science can be expressed in “perfect language,” is broadly accessible and understandable, and is therefore easy to master and replicate. We have seen why that is not true. (...)
Scientific knowledge is in fact local, person specific, private, and noncumulative. Because tacit knowledge is transmitted from person to person and contained in various reservoirs, there are greater barriers to the spread of expertise than the traditional view might suggest. Thus, the likelihood that an untrained individual with minimal theoretical knowledge could produce a biological weapon (...) is very slim.
knowledge is composed of two types: tacit and explicit.
Explicit knowledge is that which can be translated into a written form or verbalized. Equations, numbers, laboratory protocols, recipes, designs, instructions, and scientific publications are good examples of explicit knowledge. Because explicit knowledge can be imprinted onto a physical support (be it paper or computer), it can be stored for long periods of time, copied, and transferred through impersonal means, such as e-mail messages or computer files.
In contrast, tacit knowledge is composed of unarticulated skills, know-how, practices, tricks of the trade, or visual and tactile cues that are more personally held. Because of its personal and intangible nature, tacit knowledge is less easily copied or stored on a physical support. Instead, it is stored in the brains of individual scientists, technicians, or engineers, which limits opportunities for its transfer. When transfer occurs, it requires direct interaction between individuals
the acquisition of that knowledge does not occur instantly; it often involves a long- term and painstaking process of learning (...) Even when tacit knowledge can be verbalized, its use can be fraught with obstacles because portions of the knowledge may continue to escape codification. As a result, its acquisition may involve a long-term trial-and-error process, during which individuals build a personal mental library of experimentations, allowing them to gradually and often unconsciously adjust the way they do things to reach success. (...)
Learning the forms of tacit knowledge that cannot be verbalized is even more complicated, because the novice cannot benefit from the conscious guidance of the teacher. In this context, the passage of the skill from one individual to another requires not only direct contact but also a lengthy apprenticeship.
The acquisition of the skill by the recipient may follow different processes: learning by emulating or learning by doing. In the first process, scientist B observes scientist A and emulates him until the skill is acquired. Emulation can result from the conscious observation and replication of a specific way of doing things. (...) Learning by doing is the process required to learn motor skills—skills dependent on body movements that can only be acquired through personal practice. A classic example of motor knowledge is riding a bike
The development of complex products—such as automobiles, aircraft, or weapons systems—requires large teams of scientists, technicians, and engineers from a wide array of disciplines and with specific skills. Such projects are usually divided into a series of stages, integrated across functions. They call on the expertise of interdisciplinary teams, in which each individual contributes his or her personal knowledge to the development of the complex product. The final product is therefore a team effort, not the result of a scientist working alone. (...)
Through their contribution to the larger technical goal, teams of individuals produce a new form of knowledge—communal knowledge—that is spread among all contributors, but no one individual possesses the whole knowledge.
Several studies have also emphasized the local character of knowledge. Because it is created within a specific environment, by a specific set of individuals, and with a specific infrastructure, knowledge cannot be easily transferred to a new location. (...) The translation process therefore often leads to a “reinvention” of the technology, because vastly different circumstances at the new location cause the product resulting from the translation to differ from the original. (...)
The local character of knowledge is also expressed in the often inaccessible or hidden idiosyncrasies of a specific laboratory. Those idiosyncrasies cannot be standardized and transferred to a new location because they reflect individual and communal knowledge along with laboratory disciplines and routines not always recognized as part of an experiment.
Specific factors are required to allow the learning and creative process to proceed. These include the state of knowledge at the starting point, the need to create a common knowledge base, the existence of a work environment allowing informal communications, and the development of trust between collaborators. (...)
One key factor in learning tacit knowledge is the recipient’s base knowledge. Individuals absorb new ideas better when they can associate them with things they already know. In other words, building on existing expertise is easier than learning something completely new. Conversely, transferring knowledge from the source to the recipient is easier when both have knowledge in common, which results in part from similar training and backgrounds. (...)
However, building this common knowledge requires more than just formal training. Several studies have shown that informal interactions within and outside an institution are essential for the transfer and learning process to occur (...) This process is especially important for novices or beginners who, because of their limited exposure to a field, need to complement their theoretical knowledge with practical expertise. (...)
producing, creating, transferring, and learning tacit skills are part of a social process in which individuals learn through exposure to more experienced colleagues—as much as they do through their own experimentation and personal practice. As a result, knowledge is dependent on connections among individuals, their histories, and their physical circumstances, making it context dependent and not always easy to use in different circumstances.
knowledge is also stored in various reservoirs. (...) Although they contain different types of knowledge, these individual reservoirs are interdependent (...) Given the interdependence of knowledge reservoirs, it is not surprising that access to information from one but not the others strongly limits the ability to replicate past work. (...)
For example, though tacit and explicit knowledge are distinct, the boundaries between them are hard to delineate, and the two are intimately connected (...) Thus, the absence of tacit knowledge can prevent the use of explicit information and make it difficult to repeat previously successful experiments—even when attempted by experienced individuals. (...)
reservoirs of tacit knowledge are essential not only to replicating past work but also to preserving a technology. Indeed, the erosion of tacit knowledge through lack of practice, reorganization, or restructuring that affects corporate culture, work organization, routines, and communities of practice can have a deleterious effect on a technology
Written documents—be they blueprints, protocols, scientific publications, or other scientific data—are the main reservoirs of explicit knowledge, as they can be easily transferred and stored for long periods of time. Written documents, however, contain only knowledge that can be codified. Much of the knowledge created during experimental work is tacit and cannot be easily captured in written form. In addition, different types of documents serve different purposes and may only contain data useful to their target audience—be it the general public, professionals, or experts in a specific field. (...)
These written documents are all incomplete reservoirs of knowledge, which explains why they require extensive interpretation, user modifications, and adaptations to fit specific needs and work conditions. The incomplete nature of documents also stems from the fact that they record knowledge at a specific point in time, whereas knowledge creation and learning are dynamic processes.
In scientific institutions and firms, individuals are the main reservoir of knowledge, acquired through their own direct experience and observations and through collaboration with colleagues. However, in complex projects, individuals are only partial reservoirs of the knowledge pertaining to a technical artifact. (...) A wider and more complete knowledge of the technical artifact is retained collectively—or shared among teams of scientists and technicians. (...)
Because collective knowledge far exceeds the capabilities of any individual, it cannot be centrally stored; instead, it is spread across the organization. Much of this knowledge is also tacit in nature, making its transfer or reproduction difficult without the intervention of the collective.
Another, often overlooked, knowledge reservoir is corporate culture. (...) Shared culture is the product of past experiences, transmitted to new generations of employees through informal discussions, training, observation, behavior, or explicit rules of conduct. Like tacit knowledge, corporate culture conveys both certain values and scientific practices.
Communities of practice are also important reservoirs of knowledge. Unlike corporate culture, which is confined by the walls of an institution, communities of practice embrace a whole professional community. Communications and exchanges between individuals within communities of practice allow methods and approaches to spread. In the process, such exchanges enrich personal and communal knowledge, as well as that of the corporate culture.
A final knowledge reservoir is the organizational structure of an institution. In complex projects, in which different teams perform specific tasks, the organizational structure reflects the sequence of tasks. More precisely, the organizational structure indicates how the specialized knowledge created within each division is transferred to and used by other divisions, creating a new institutional knowledge.
When employees retire or organizations are restructured, mistakes, disruptions in production, or decreases in output quality often follow, due to a loss of know-how (...) knowledge needs to be replenished continuously; otherwise, it depreciates. Scholars in the science and technology field have also observed that tacit knowledge can decay over time if it is not used, practiced, or transmitted to a new generation of specialists. Trying to re-create lost knowledge can be difficult, if not impossible.
Knowledge loss can occur at different levels. (...) At the organizational level, a massive loss of personnel due to restructuring or retirements may cause an institution to lose valuable experimental knowledge and insights. (...) Knowledge loss may also occur at the division or unit level when people fail to document modifications in processes or mistakes in engineering designs, leaving newcomers to face insurmountable hurdles. (...) Finally, knowledge loss may occur at the individual level, when a person stops practicing an activity.
these changes have had a profound impact on the various knowledge reservoirs that constitute the nuclear complex. Personnel downsizing affects communal knowledge and the communities of practice that were formed at the laboratories during fifty years of nuclear weapons development. The design and testing of nuclear weapons brought together a varied group of individuals, including physicists, chemists, engineers, technicians, and explosives specialists, who created an identity around the task of transforming a physics principle into a complex working weapon. Their knowledge was passed on from one generation to another through formal training, informal communication, practice, apprenticeship, and gradual inclusion of novices into this community of practice. Testing of nuclear weapons created a common knowledge in the form of explicit data, but also tacit practices and values.
This communal knowledge started to erode with the departure of individuals who contributed to nuclear weapons design, production, and testing (...) Those who remain have their own personal knowledge, but due to the complexity of the nuclear enterprise, no individual fully understands all the details of nuclear weapons design, development, production, and testing. (...) As a result, the nuclear complex is leaking knowledge from all reservoirs.
the Soviet bioweapons program essentially failed to produce the new generation of bioweapons it set out to develop in the 1970s despite a prodigious investment in human, material, and financial resources. The U.S. bioweapons program also was largely unable to produce a weapon that met military requirements, while more recent state and terrorist programs have failed at various or all stages of bioweapons development in spite of having access to the material and financial resources.
Given the fragility and unpredictability of bio-agents, and the challenges of tacit knowledge creation and institutionalization, the environment required to achieve fruitful [learning] interactions does not occur spontaneously; it must be engineered in an especially appropriate organizational context. (...)
Organizational and managerial factors thus constitute key determinants of success because they have a direct impact on knowledge creation and transfer within a program. Evidence from past state and terrorist bioweapons programs also underscores the importance of four additional factors that affect successful systems integration: political support or intrusion, the overall program’s economic circumstances, program location, and the type and timing of foreign technical assistance.
The extant literature has identified two key organizational factors that promote the efficient use and transfer of knowledge: first, a structural organization that ensures personnel proximity and mobility; and second, the deployment of integrative mechanisms that ensure the coordination and synchronization of tasks and stages.
Given that the acquisition and use of tacit skills and know-how requires direct interactions between people, often over a prolonged period of time, proximity between individuals and mobility within their organization are essential vehicles for knowledge transfer. This is especially true in bioweapons programs, which must deal with unpredictable microorganisms, novel problems, and complex uncertainties, all of which demand close and frequent collaboration and coordination between diverse technical staff. (...)
Proximity therefore plays a crucial role in creating connections between different reservoirs of knowledge and enriching their component parts. It fosters communities of practice, in which direct communications between members serve as conduits for the transmission of values, practices, and certain ways of doing things. Proximity also promotes the likelihood of unplanned interactions between people, which can improve coordination between functions within an organization. At the individual level, proximity allows greater feedback (...) Conversely, distance impairs the transfer of knowledge, because it limits opportunities for knowledge spillover through informal and serendipitous interactions between individuals. (...)
Another key requirement of knowledge transfer is personnel mobility. Allowing people to move between divisions, or from one location to another, can not only reinforce the benefits of proximity but also compensate for the lack of physical proximity.
In the bioweapons field, the importance of stage coordination is illustrated by the challenge of scaling up fragile biological organisms. Schematically, the development of a bioweapon proceeds through five main stages.
First, during the research phase, teams of scientists and laboratory personnel who specialize in bacteria, viruses, or toxins study and develop an agent with appropriate characteristics (virulence, antibiotic resistance, pathogenicity) for weapons use.
Then another team that specializes in production process development takes over the sample to design ways to produce the agent in slightly larger quantities while maintaining all the agent’s desirable qualities.
The product is then moved to the production phase, where engineers and technicians test the production process by producing small quantities of the agent in a pilot plant. The objective here is to identify potential problems that might prevent a further scale-up to industrial quantities.
In the fourth stage, animal experts test the resulting product on primates or other laboratory animals to model its effects and determine whether it can produce the expected results on humans.
The last stage, weaponization, is executed by engineers, explosives experts, mathematicians, and statisticians who design delivery mechanisms and test the dispersion of the agent in a specific weapon delivery system.
Although in theory these stages occur sequentially, in practice, the evolution from laboratory sample to manufactured product is never a linear process (...) The passage from one stage of product development to the next requires a constant back-and-forth between stages and direct interaction among participants. This is not easily achieved when the programs include hundreds or even thousands of individuals located at different sites.
Because large and complex projects are composed of many moving parts, integrating these parts into a coherent whole—a process known as systems engineering—is particularly challenging and affects a project’s success or failure (...) systems engineers are individuals who see a project through, from its inception to its end, and synchronize and align the work of the different participants. Thus, they contribute not only to the development of the project concept but also to its design, research and development, and execution.
new knowledge is sometimes rejected due to trust issues or lack of common technical language. Managers have an important role to play in ensuring acceptance and use of new knowledge, and their influence manifests itself in three important areas.
First, they are essential in establishing a social context that promotes trust and cognitive cohesion—the creation of a common technical language—among personnel, which increases the chances that personnel will accept and effectively use their colleagues’ knowledge. Second, they play a particularly important role in creating the proper environment for knowledge institutionalization (...) Finally, managers are essential conduits for allowing sustained communications between political decision makers and program implementers, to ensure continued political and financial support for the program. (...)
Managers can promote organizational learning by creating a corporate culture that promotes openness and rewards information sharing. One strategy for creating such an environment involves implementing a shared vision of what should be accomplished. (...)
Conversely, the emphasis on negative incentives, in the form of punitive actions for errors, can lead to avoidance strategies, preventing the identification of problems and their resolution. (...) sanctions for failure were stiff in the Soviet and Iraqi [bioweapons] programs as well as in the Japanese terrorist group Aum Shinrikyo, leading personnel to report fake results or even fabricate data to please higher-ups and avoid punishment.
Complex projects such as weapons programs rely on heterogeneous groups of individuals with different competencies, who frame and approach problems differently. This creates cognitive boundaries and trust issues that are difficult to overcome, making the knowledge created by one unit not readily usable by another unit. Weapons programs also must contend with cultural differences between military and civilian personnel, who operate according to different cultural and administrative rules. Work on a common project therefore requires the creation of a common technical language that will allow diverse technical communities to collaborate, learn from, and trust one another (...)
informal ties created through socialization lead individuals to create friendships that increase their willingness to share information and help each other. The stronger the ties between individuals, the more time and effort they are willing to invest in providing assistance to each other and transferring their knowledge. Friendship and frequent communications also play an important role in the development of trust, an essential ingredient in knowledge sharing and use.
In order for an institution to make progress and to innovate, it must ensure that knowledge acquired by an individual or a group can spread across the organization. This not only allows others to benefit from that knowledge but also permits the whole organization to adapt to new knowledge by, for example, adopting new routines or modifying existing production processes. (...) Openness to the outside world is another means of enhancing organizational learning. (...)
Finally, nurturing openness and an environment where people can safely identify problems and search for solutions also requires communication channels that easily pass information up and down the hierarchy. This permits knowledge created at various work levels to spread across the entire institution, if needed.
Achieving proper conditions for knowledge institutionalization in weapons programs is particularly challenging because they are subject to secrecy requirements, which translates into restricted access to information through compartmentalization. In many cases, too, weapons programs are placed under military control, which tends to favor strict vertical hierarchies, centralized decision making, and bureaucratic rigidity—conditions that are poles apart from the required openness and flexibility needed for knowledge institutionalization. The challenges are even steeper in illicit bioweapons programs launched or maintained after the signature of the Biological Weapons Convention, because maintaining absolute covertness places greater demands on management, making the creation and transfer of knowledge problematic. (...)
developing bioweapons while operating under the constraints of covertness not only imposes additional costs to create the protective layers necessary to prevent detection but also complicates the already stiff challenges of knowledge management. (...) Soviet authorities decided against creating lateral linkages within their bioweapons program to limit individual knowledge of the program. This caused many scale-up nightmares
several external factors can also affect scientific outcomes because they can interfere with the stability of the work environment, the continuity of scientific work, and the integration of its constituting parts. Stability in the work environment is essential for ensuring the proper use and transfer of knowledge and allowing personnel to accumulate knowledge over time. (...)
In past bioweapons programs, four exogenous factors have conspired to upset a program’s stability, continuity, and integration: (1) the level of political priority and intrusion in a program, sometimes teamed with a lack of cohesion in decision making; (2) the economic circumstances of a program; (3) the location of a program; and (4) inappropriate or untimely foreign technical assistance.
when political elites interfere with scientific decisions, program delays and failures occur. On the other hand, when scientists are given the opportunity to develop a professional culture, with minimum interference from political leaders, program success is more likely (...)
In these three cases [the Iraqi, Soviet, and Aum Shinrikyo biological weapons programs], interference took the form of goal setting that was out of sync with what scientists could actually do; placement and replacement of scientists based on political loyalty rather than competence; and interference with the natural flow of scientific discovery, requiring scientists to skip essential phases to reach the requested result faster. In all three cases, such intrusions resulted in multiple disruptions in scientific work, the emergence of questionable scientific behaviors—faking or fabricating results—in some cases due to the creation of a cadre of incompetent scientists; and a loss of knowledge, which eventually negatively affected these programs’ paces and outcomes.
Typically, several government agencies and military services are involved in decision making for a program. These actors often have different and sometimes conflicting views about a program’s goals and priorities. Reconciling these diverging views can be a daunting task, particularly when funding decisions are not centralized. (...)
About a dozen agencies were involved in decision making for the U.S. bioweapons program. Such dispersion of responsibility and decision making not only nurtured conflicts and disagreements about the program’s objectives but also negatively affected the program’s direction and scientific output.
The location of a program also becomes an important factor affecting program success or failure. Several studies have shown that location is important because the properties of the materials, equipment, or parts necessary for scientific work may differ from one location to another. Variable properties can negatively affect replication of past work and the transfer of technologies to a new site, ultimately complicating integration of parts and stages. (...) Minute changes in material or component characteristics, due to the use of different suppliers at a new location, can prevent successful integration and derail an experiment. (...)
The role of location and the properties of materials used in scientific work are nowhere more important than in the biological sciences, where scientists work with live agents that are unpredictably sensitive to environmental conditions. For example, the properties of reagents and other materials used in scientific experiments may differ from one location to another and frequently vary seasonally. An experiment conducted successfully in one location may not be reproducible in another because of the varying properties of the materials used, even when the same individual conducts the experiment.
Appreciating the uneven results of foreign technical assistance on weapons development does not mean that foreign assistance is unimportant. Indeed, it is a central feature of eventual success. All programs have thus far had to elicit the help of friendly countries simply because the breadth of resources required to bring a program to fruition usually far exceeds the capabilities of an individual state. Even the two superpowers during the Cold War could not claim that their nuclear and biological weapons programs were fully endogenous. (...)
Consequently, evaluating the role of foreign assistance in weapons development requires digging deeply into the context within which it is offered. Practically, this requires asking two important questions. First, what is the recipient program’s absorptive and integrative capacity? Second, do the timing and nature of the assistance coincide with the recipient’s absorptive capacity? If a recipient program does not have sufficient knowledge and industrial bases to use, adapt, and integrate the received assistance into their own circumstances, it is likely that the assistance will, at the very least, remain unused, or it may effectively delay and possibly impede progress altogether.
Therefore, a weapons program should be viewed as an integrated system: it requires specific and essential parts that work individually, but these parts also need to operate when assembled as a whole. Here lies one of the key shortcomings of current threat assessments. Currently, emphasis is placed on only a few variables—material resources—at the expense of other essential ones. In addition, analysis tends to be based on the inventory of individual parts rather than on how they operate together. This shortcoming is particularly damaging in bioweapons threat assessments: due to the fundamental unpredictability of bio-agents and the resulting increased need for integration of teams and stages, bioweapons programs are especially sensitive to the variables that negatively affect integration. This may account for the limited successes achieved in both large-and small-scale bioweapons programs (...)
A bioweapons program cannot progress rapidly if its interdependent stages and functions are not coordinated and integrated. The division of labor resulting from the involvement of different disciplines and skills inevitably fosters barriers between people. Organizational division and work structure can also foster additional impediments, especially if a program is more concerned with ensuring covertness than creating the conditions for knowledge production and transfer.
the case studies underscore how difficult it is to produce working bioweapons, however vast and resourceful the program. The fragility and unpredictability of microorganisms require that state and nonstate actors create meticulous organization, management, and sustained coordination of all the actors affecting a program over time, not only to permit success in the laboratory but also to transform a laboratory sample into an agent that can survive scale-up, weaponization, and delivery. (...) These ideal conditions are difficult to achieve under any circumstances, let alone those surrounding the constraints of maintaining covertness—particularly under autocratic and violent regimes.
These findings have implications at four important levels.
- First, they have an impact on national and international security by suggesting a new narrative about bioweapons developments that has greater dissuasive power to those who might seek this technology than does the current technology-based discourse, which emphasizes ease of development and accessibility of resources.
- Second, they underscore the importance of strengthening the Biological Weapons Convention (BWC) and call for a review of current nonproliferation and counterproliferation policies.
- Third, they offer new insight on how to achieve more accurate threat assessments of suspect state or terrorist programs and design more tailored responses to these threats.
- Finally, they question the design of current policies aimed at limiting the spread of knowledge as well as the widely shared belief that new technologies will erase the hurdles associated with bioweapons developments.
Since 2001, the United States has spent more than $60 billion on biodefense programs designed to detect the release of harmful agents (the BioWatch program), produce medical counter-measures (the Bioshield program), create stockpiles of drugs and vaccines, increase research on a set of agents deemed most likely to be used in a bio-attack (the so-called Select Agent List), and strengthen laboratory security. The Bioshield and BioWatch programs have been harshly criticized: after a decade of effort, the former did not produce any new vaccines, while the latter has experienced repeated false alarms in its network of biosensors deployed in major cities, and still requires the physical removal of filters for analysis in the laboratory, which can take several days.
the current biothreat narrative and its policies, with their highly publicized shortcoming and failures, have reinforced the belief that the United States and its allies are not prepared to respond to a bio-attack in spite of a significant investment in funds and efforts over the past decade, and therefore remain very much vulnerable to such attacks. Rather than dissuading an adversary, this narrative only reinforces the desirability of developing bioweapons
dissuasion is defined as taking action to decrease the benefits or increase the cost of such weapons, thereby creating strong barriers to entry. Regarding bioweapons, none of these objectives have been achieved: the current biothreat narrative presents a cost/benefit ratio in favor of developing bioweapons, and its concomitant policies have not addressed the real barriers to entry in the bioweapons field—the acquisition of expertise and its determinants. (...)
The key message to convey is that bioweapons developments are not only difficult but highly uncertain. The fragility and unpredictability of living microorganisms creates a natural barrier that is challenging to overcome without the appropriate expertise (...)
The acquisition of expertise is therefore the key barrier, which is overcome not simply by means of acquiring scientific documents but through prolonged hands-on experimentation, requiring the cooperative work of a community of experts open to the outside world, who have expertise adapted to the types of agents under study and the skills required to move a natural agent through the various stages of research, development, production, and weaponization. Creating such a knowledge base can take decades and ultimately still fail (...)
The third hurdle lies in creating the proper organizational, managerial, and social conditions that foster knowledge creation, transfer, and institutionalization while also managing external factors
One of the gravest consequences of the biothreat narrative has been to focus attention and nonproliferation policies almost exclusively on frustrating material procurement. To be sure, these policies, including export control and UN Resolution 1540, are essential elements of nonproliferation and should be continued. However, material resources are not the key barriers to bioweapons development. It is important, therefore, to reorient current and future policies toward preventing the acquisition of expertise. Since proximity, direct contact between scientists, cooperation, and the stability and continuity of scientific work are essential to scientific progress, nonproliferation and counterproliferation efforts should focus on disrupting these factors.
The BWC is an essential nonproliferation tool, not only because of its undeniable normative value but also because it can be used to directly impede knowledge acquisition. But to fully use the nonproliferation power of the BWC, the international community must revive the idea of a formal verification mechanism to the treaty. (...)
To be sure, the verification measures as proposed in 2001 would not have caught a proliferator red-handed. But the value lies in their ability to disrupt the continuity of bioweapons work and delay progress. (...) Thus, interruptions caused by inspections, or their expectation, can set back ongoing experiments or development work, particularly when they occur at sensitive stages (...) Changes to experimental protocols are done on the fly, and if they are not recorded, they may be lost in the confusion of a move, a temporary interruption, cleanup, or the hiding of documents and materials, thus preventing staff members from resuming work where they had left off. If the interruption is long lasting, knowledge can be lost for good
Disruptions of a program’s scientific stability and continuity can be achieved in several ways. (...) the threat of military or police operations in nearby areas can keep the suspect country or group permanently fearful of detection, obliging the program to move repeatedly, conceal its activities, or stop work entirely, if only temporarily.
Second, sabotage, a strategy used in the past to target equipment, can also effectively disrupt scientific work (...) Sabotage can be particularly damaging in the bioweapons field to the extent that it affects production and scale-up equipment. Equipment malfunction during these stages not only halts the accumulation of knowledge when theoretical concepts are tested during production but also could destroy a batch of bacteria or viruses undergoing production. Both outcomes would result in major program delays.
Disrupting team composition can also produce program delays and knowledge loss. Because teams take a long time to jell and work effectively, forcing changes in their composition can disrupt cohesiveness and delay work, as well as foment distrust among members, particularly in the context of a hostile environment. (...)
Playing on a program’s fear of detection or a mole within the organization can also constitute an efficient strategy to push the program toward an organizational model that limits or altogether prevents communications between knowledge reservoirs
Compelling the suspect program to locate the various stages of a bioweapon’s life cycle at different sites and limiting connections between them can further reinforce organizational impediments. The threat of police or military activity and the possibility of international inspections can be equally effective at compelling the suspect group or country to retreat to an adverse organizational model that shatters a program into pieces, spreading it to different locations while compartmentalizing its component parts. With reduced direct interactions, the transfer of tacit knowledge and task coordination are compromised.
The starting point of any bioweapons threat assessment should be an evaluation of the potential for a state or terrorist group to manipulate and process successfully fragile microorganisms. To that end, data collection and analysis efforts should first of all attempt to determine whether the individuals involved have sufficient knowledge.
roughly three categories of individuals may be involved in a nascent bioweapons program: the novice, the sub-expert, and the expert.
- A novice is a person with a generic educational background in the discipline but no specific or practical expertise. Aum Shinrikyo’s program was staffed with novices. (...)
- The sub-expert is a person with a specific expertise—say virology or bacteriology—and advanced theoretical knowledge but no practical expertise in developing biological weapons. Most scientists in the Iraqi and South African programs belonged to this category. (...)
- The expert is an individual with advanced theoretical knowledge and practical expertise in a domain directly applicable to bioweapons work, and possibly even bioweapons expertise. The American and Soviet programs both included this category of scientists. (...)
These distinctions are important because they reveal how different people learn and use their knowledge—that is, what their absorptive capacity is and, consequently, how quickly their work progresses (...) The level of expertise also determines a scientist’s/technician’s ability to use outside assistance or technology and adapt it within his or her own context.
because expertise in one domain does not necessarily transfer to another domain, it is important to investigate whether individuals working in a bioweapons program have expertise that corresponds to the actual work they are doing, and whether the program has access to the whole gamut of knowledge required to ensure passage from one stage to the next in a bioweapons life cycle. (...)
Therefore, the further removed the resident expertise is from the work being conducted, the more difficult and lengthy is the acquisition of knowledge. In this context, even experts in one discipline might have to take the same learning path as sub-experts or novices if their expertise does not fit the work being conducted. (...)
If one has familiarity with one type of biological organism, this can help facilitate some aspects of the learning process, but it will not eliminate the need for trial-and-error experimentation on the new organism.
Brain drain prevention programs were launched in the 1990s as a result of the breakup of the Soviet Union, which unleashed new threats of proliferation of nuclear, chemical, and biological technology, material, and expertise. Spearheaded by the United States, the international community designed and implemented assistance programs to secure weapons facilities and their dangerous material, and offer jobs to weapons scientists to prevent them from offering their services to other states or terrorist groups. The Department of Defense Cooperative Threat Reduction (CTR) Program was and probably remains the largest nonproliferation program in the Soviet Union. Although it first focused on nuclear and chemical proliferation, it expanded its reach to include bioweapons facilities and scientists in the late 1990s. The program has been quite successful at securing bioweapons facilities and their collections of pathogens, as well as improving the safety of laboratory work at these facilities.
On the brain drain front, however, there remain reasons for concern about the effectiveness of these nonproliferation activities, primarily due to the way brain drain programs are designed and implemented. (...) Three design flaws have marred the effectiveness of the CTR Program.
First, most of the research projects funded under the CTR Program maintain bioweapons scientists at their former facilities, where they work on biodefense-oriented projects that involve many of the same dangerous pathogens that former bioweapons scientists worked on in Soviet times. In addition, these scientists often work with the same Soviet-era colleagues. These two factors allow them to maintain many of the tacit skills and communal knowledge developed under the Soviet bioweapons program instead of facilitating its erosion. (...)
Second, the CTR Program is facility based, meaning it only supports scientists working at certain bioweapons facilities. Once these scientists leave their former institutes (...) they cease to qualify for this support, even though a covert program could advantageously use their expertise. The facility-based approach also has the major disadvantage of not distinguishing between scientists and technicians who might actually pose a threat and those whose expertise would not benefit a covert program. (...)
Finally, the program focuses on facilities previously believed to be part of the core Soviet bioweapons infrastructure, at the risk of neglecting facilities located in the other two circles, such as the anti-plague system. Anti-plague scientists may pose as much of a threat as those employed in core facilities
the CTR Program should modify its current approach, which emphasizes the expansion of the program to as many facilities as possible, to one that engages the facilities and individuals that pose the greatest threat. (...) In general, the CTR Program should de-emphasize biodefense projects and create incentives for scientists to exit the bioweapons field, in order to reduce the overall number of individuals able to maintain their bioweapons-specific skills.
written documents are decidedly incomplete reservoirs of knowledge that rarely allow replication of past work without the intervention of the original authors, particularly when the work is conducted at a different location and with different material and equipment. (...)
Therefore, restricting scientific publication does not support, but actually works against, nonproliferation goals. It also perpetuates the false belief that the fast pace of new scientific developments, aided by a so-called revolution in biotechnology, will eclipse dependence of locally based scientific skills. (...) Apart from the obvious negative consequence for public health—restricting the spread of scientific data can impede public health authorities’ ability to prepare against and respond to a localized outbreak or pandemic—publication restrictions fuel suspicions that the United States is using this data for military purposes.
new technologies do not, by themselves, allow replication of past work, or transform an untrained individual into a bioterrorist or bioweaponeer overnight. In fact, empirical studies and interviews with bench scientists indicate that new technologies are not necessarily easy to use, even by trained scientists. They require their users not only to possess prior base knowledge but also to acquire new expertise to solve novel problems created by the use of the technology. (...) Thus, even new-generation technologies require prior expertise and the acquisition of new skills to interpret and troubleshoot problems that are certain to arise. (...)
In the bioweapons field, unless future technologies can render biomaterials behavior predictable and controllable, and allow scientists to transition easily from work with one agent to another, the role of expertise and its socio-organizational context will remain critically important barriers to bioweapons development.
Making contributions to the development and use of a bioweapons program a crime against humanity could serve as a stronger deterrent. Criminalization of bioweapons development at the national level is a requirement under UN Resolution 1540, and several countries have already enacted legislation making it a crime for terrorists to develop, use, or acquire mass destruction weapons, including biological weapons. (...)
However, both the BWC and UN Resolution 1540 suffer from insufficient implementation. For example, the content of the legislation reported by individual countries to the UN Resolution 1540 Committee is unequal, as is enforcement of the legislation at the national level. Yet 1540 Committee personnel do not have the mandate to evaluate this legislation or issue a template of what constitutes exemplary legislation.
I read chapters 1-3 and 7 but not chapters 4-6, which respectively deal with the Soviet, American, and several small-scale biological weapons programs.
This is really helpful, thanks for sharing. I had heard the "tacit knowledge is important/hard" category of argument before, but this was unusually well articulated.