NCW, C4ISR, IO and RMA: Toward a Revolution in Military Intelligence? No military forces ever have placed such faith in intelligence as American onesdo today. Advocates of the “revolution in military affairs” ( RMA ) assume thatinformation ( as technology or superhighway or revolution or age ) would transform theknowledge available to armed forces, and thus their nature and that of war. This faith iscentral to American doctrine and policy. Joint Visions 2010 and 2020, which guidestrategic policy, predict forces with “dominant battlespace awareness” , better knowledgethan and a “frictional imbalance” and “decision superiority” over an enemy, andunprecedented flexibility of command: the ability to combine freedom for units withpower for the top, and to pursue “parallel, not sequential planning and real-time, notprearranged, decisionmaking”. 1 Officials have created new concepts about intelligenceand command, aiming to pursue power by fusing, into systems, matters which once weresplit into “stovepipes”, and new forms of information technology. These concepts includenet-centric warfare ( NCW) , the idea that armed forces will adopt flat structures, workingin nets on the net, with data processing systems at home serving as staff for the sharp endthrough reachback; C4ISR ( command, control, communications, computers, intelligence,surveillance and reconnaissance; loosely speaking, how armed forces gather, interpretand act on information ); the “infosphere”, the body of information surrounding anyevent; and “IO” (Information Operations), the actions of secret agencies. Progressives and revolutionaries ( conservatives need not apply) debate theseissues in detail and in principle. The Marine Corp’s draft doctrine on IO denies thattechnology can solve all problems, and defends “our timeless fighting principles”. Armydoctrine too gives C4ISR a Clausewitzian cast. It can “reduce the friction caused by the fogof war” and help impose one’s will on the enemy, but “achieving accurate situationalunderstanding depends at least as much on human judgment as on machine-processed information—particularly when assessing enemy intent and combat power ...Uncertainty and risk are inherent 2in all military operations”. Conversely, the Pentagon’s Director of Force Transformation,Admiral Cebrowski, claims that all principles of classical strategy, like “the relativelyhigh value accorded mass”, stemmed from “the dearth of timely, accurate, andcomprehensive information. In the relative absence of information, mass provided theinsurance against what Clausewitz called the fog of war”. These problems had ended;classical strategy was dead; a “new theory of war based on information age principles andphenomena” was needed. Revolutionaries assume C4ISR will function in a systemprecisely as a person sees the world, turns data to knowledge and acts on it. They believearmed forces can comprehend an enemy and a battle perfectly, and act without friction.David Alperts, a founder of NCW, holds, we will effectively move from a situation in which we are preoccupied with reducing the fog of war to the extent possible and with designing approaches needed to accommodate any residual fog that exists to a situation in which we are1 “Joint Vision 2010” and “Joint Vision 2020”, www.dtic.mil/doctrine ; for a bibliography, cf. John Ferris,“The Biggest Force Multiplier? Knowledge, Information and Warfare in the 21st Century”, in Alistair Dallyand Rosalind Bourke (eds), Conflict, The State and Aerospace Power ( Canberra, 2003), pp. 149-165.2 Marine Corps Combat Development Command, Draft “Information Operations”, 25.9.01, USMC,Doctrine Division Home Page; FM 100-6, 27.8.96, “Information Operations”, www.adtdl.army.milcgi-bin/adtl.dll.; FM 3-0, Operations, 14.6.01, 5.75, 6.38, 11-47.
preoccupied with optimizing a response to a particular situation...we will move from a situation in which decision making takes place under “uncertainty” or in the presence of incomplete and erroneously sic information, to a situation in which decisions are made with near “perfect” information. 3 All sides in this debate assume intelligence will have great power. They take itstriumphs for its norms. Thus, in 1995, the USAF chief, General Fogleman, discussingULTRA and FORTITUDE, said, “Throughout history, soldiers, sailors, Marines andairmen have learned one valuable lesson. If you can analyze, act and assess faster thanyour opponent, you will win!”—unless, of course, it is stronger or smarter or luckier thanyou. 4 The military exponents of Information Warfare ( IW) assign unprecedented weightto intelligence in war. In 1995 Colonel John Warden, USAF planner and theorist ofairpower, held “Information will become a prominent, if not predominant, part of war tothe extent that whole wars may well revolve around seizing or manipulating the enemy’sdatasphere”. George Stein wrote “Information warfare, in its essence, is about ideas andepistemology—big words meaning that information warfare is about the way humansthink and, more important they way humans make decisions...It is about influencinghuman beings and the decisions they make.”5 Faith in intelligence and IO underlieCommand and Control Warfare ( C2W), the main form of operations the United Statesplans to fight, a version of blitzkrieg which seeks “to deny information to, influence,degrade, or destroy” the enemy’s “information dependent process”, so to shatter itsability to perceive and command. 6 Revolutionaries advocate a higher mode of war,Rapid Decisive Operations ( RDO), which features “High Quality Shared Awareness”,“Dynamic Self-Coordination”, “Dispersed Forces”, “De-massed Forces”, “DeepSensor Reach”, “Compressed Operations and Levels of War”, “Rapid Speed ofCommand” and the need to “Alter Initial Conditions at Increased Rates of Change”,by exploiting all these principles “to enable the joint force, across the cognitive,information and physical domains of warfare, to swiftly identify, adapt to and change anopponent’s operating context to our advantage”. RDO will open with the pursuit of a“Superior Information Position ( Fight First for Information Superiority)” andbecome “knowledge-centric”: The creation and sharing of superior knowledge are critical to RDO. ...Decision makers, enabled by study, judgment, and experience, convert information into knowledge and situational understanding, which is the key to decision superiority – the ability to make better decisions faster than the adversary... IO are the3 Admiral Cebrowski, “The Small, the Fast, and the Many”, NetDefense, 15.1.04, p. 10; David Alperts,“The Future of Command and Control with DBK”, in Stuart E. Johnson (ed), Dominant BattlespaceKnowledge, NDU Press, October 1995. For a different reading of the relationship between intelligence andstrategy, cf. Ferris, J.R., and Michael Handel, "Clausewitz, Intelligence, Uncertainty and the Art ofCommand in Modern War", Intelligence and National Security, vol. 10, no. 1, January 1995, pp. 1-58.4 General Ronald Fogleman, speeches, 25.4.1995, “Information Operations: The Fifth Dimension ofWarfare”, Defense Issues, Volume 10, Number 47, www.defenselink.mil/speeches/1995 and“Fundamentals of Information Warfare-An Airman’s View”, 16.5.1995, www.af.mil/news/speech/current.5 George Stein, “Information Warfare”, Airpower Journal, Spring 1995; Colonel John A. Warden III, “AirTheory for the Twenty-first Century”, Aerospace Power Chronicles, “Battlefield of the Future: 21stCentury Warfare Issues”, 1995.6 Joint Pub 3-13.1, Joint Doctrine for Command and Control Warfare ( C2W), Joint Chiefs of Staff,7.2.96
information equivalent of maneuver and fires... In planning for effects-based operations, knowledge is paramount. 7 The premier US military exercise of 2002, Millenium Challenge 02, tested everytransformed force and component of RDO, particularly “information/ knowledgesuperiority” and the establishment of “a knowledge network”. The USAF commandernoted that the aim was “machine-to-machine talk”, so that a commander can “create anair tasking order with one push of a button. I can see the entire battle in a way that ifthere’s something I don’t like, I can fix it”. The “Unified Quest 03” and “Quantum Leap”wargames of 2003 addressed such issues in more detail, while advocates pushed strategyfrom the age of Clausewitz to that of the Borg. 8 Cebrowski held that power soon wouldbe defined by “’information fraction’...the measure of a system’s ability to access andcontribute to a larger information network” through NCW and C4ISR. This is the age of the small, the fast, and themany. Small: Power and size are uncoupled. Fast: A shorter response with a faster rise time more precisely placed in time and space. Many: The power of the collective at lower cost over a larger area. Rebalance for the information age. “Demassification” through increasedinformation fractions. Simplification through adaptive relocation ofcomplexity & the human. Networked components vice integratedsystems. Operations based on assured access, information superiority, control of initial conditions and rates of change. A priori access to the domains of conflict. Secure a superior information position and convert it to a competitive advantage. Leverage the path dependency of conflict. 9 The revolutionaries conceptualise war as game and strategy as shooting. Theyassume that to be seen is to be shot, to be shot is to be killed, and to be fast is to win.Warden defines “a very simple rule for how to go about producing the effect: do it veryfast...the essence of success in future war will certainly be to make everything happenyou want to happen in a very short period of time—instantly if possible”. 10 Thesetendencies are reinforced by the use of Colonel John Boyd’s OODA cycle—Observe,Orient, Decide, Act—to describe all conflict on all levels of war, with the aim usually7 Department of Defense, Transformation Planning Guidance, April 2003, APP 4, “Joint ConceptGuidance, www.oft.osd.mil/; U.S. Joint Forces Command, A Concept for Rapid Decisive Operations,RDO Whitepaper, J9 Joint Futures Lab, 2.3, 4.1, 184.108.40.206 USJFCOM, Millenium Challenge 02, www.jfcom.mil/about/experiments/mc02 ; 9.8.02, “Air Forceportion of experiment ends with positive result”, www.jfcom.mil/newslink/story/archive/2002/no080902;“Unified Quest 03” www.jfcom.mil/about/experiments/uq03 .9 Admiral Cebrowski, “The Small, the Fast, and the Many”, NetDefense, 15.1.04, p. 1010 Warden, “Air Theory for the Twenty-first Century”, ibid
defined as moving through the cycle faster than one’s opponent: wiser heads urge thatthis edge be used to think more rather than simply act faster. That model, derived fromBoyd’s reflections on his experience as a fighter pilot in the Korean War, is a good meansto conceptualise one on one combat. It is less useful for war. In a boxing match, speedmay equal victory; in strategy, cries of “faster! harder!” produce premature ejaculation.Focus on the OODA cycle, “sensors to shooters”, “one shot one kill” weapons and theidea that armed forces can act almost without friction on near perfect knowledge, has ledto a fetishisation of speed and the tacticisation of strategy. These ideas frame those about intelligence. The assumption is that intelligencewill be an engine fit for a fine tuned, high performance, machine--reliable, understood,useful, usable, on-call. One can learn exactly what one wants to know when one needs todo so, and verify its accuracy with certainty and speed. The truth and only the truth canbe known. It will show what should be done and what will happen if one does. Actiontaken on knowledge will have the effect one intends, nothing more or less. Intelligenceexperts in the military-academic complex have attacked these ideas. Williamson Murraynotes that a key to RDO, the idea that “Operational Net Assessment” ( ONA) will turnknowledge to power by constantly updating and fusing all data on everything related to awar, ignored every known problem in intelligence. At a strategic level, net assessmentrarely did more than bean counting. It usually fell victim to worst or best caseassessments and mirror-imaging. Achieving ONA would require “a revolution in theculture of intelligence”, a “move away from the search for the predictive to an emphasison a broader, intuitive understanding of potential opponents”, from focus on collectionand technology toward foreign languages, culture and history. Michael Handel arguedthat intelligence, once undervalued, had become oversold. “If it sounds too good to betrue”, he briefed officers, “maybe it is”. 11 American doctrine about intelligence andoperations, too, treats the relationship between these matters carefully and well. 12 The Pentagon expects normal intelligence to be as good as it ever has been, morecentral to planning and operations, and to be transformed along with every other elementof power. In January 2003, the Chairman of the Joint Chiefs of Staff, General Myersnoted, “we have always tended to have this situation where intelligence people are in onestove pipe and the operators in another and we are real happy if they talk together. Whentoday’s world requires that they be totally integrated...you can’t have an intel pod, throwit over a transom to an operator and say, here’s what we know. This has got to becontinuous, 24/7 sort of relationship and synergistic to the point where operations helpwith intel and vice versa”. 13 Commenting on Operation Iraqi Freedom, Cebrowski noted the intelligence analysis problem, where we have all of these intelligence sources, and they all produced their products and reports and fed databases, all of which11 Williamson Murray, “Transformation: Volume II”, pp. 10-17, in Williamson Murray (ed),Transformation Concepts for National Security in the 21st Century, Strategic Studies Institute, The USArmy War College, September 2002; an untitled briefing paper by Michael Handel, copy in my possession,circa 2000. These papers offer fundamental critiques of assumptions about intelligence in the RMA; cf,Ferris, “The Biggest Force Multiplier?”12 cf. n. 2, 6 and 34.13 Defense Writer’s Group, 22.1.03, interview with General Richard Myers,www.oft.osd.military/library.cfm?libcol=
are stove-piped. The analysis functions are similarly stove-piped. Essentially, we have an intelligence community that is organized by wavelength. But it needn’t be that way. It could be more like this, where your intelligence is organized around the demand functions of warning, force protection, and warfighting intelligence, where you have data mediation layers that are able to pull together all source information, plot it geo-spatially, and generate the kinds of displays in which a senior leader’s question can in fact be answered at very, very high speed. 14He advocated “a new demand centered intelligence system”. Since 1970, Cebrowski held,intelligence had become unbalanced as the quantity of information rose exponentially, thepower of analysis grew in a linear fashion, while specialists alone collected, processed,(and hoarded ) material from each source. This caused overload and strangulation ininformation. Too much data was available, too little of it used and even less coordinated,because it was divided into watertight pots defined not by function but source. Never wasall the data on any topic brought together. Agencies collected what they did because thatwas what they did; the customer was forgotten. So to solve these problems, material fromall sources should go to an “Information Dominance Center“ ( IDC) for analysischaracterized by “Continuous merge ...Megadata, All Source, Open Source andGeospatial Data, Dynamic Collection, Visualization”. This idea, like that of ONA,assumes analysts constantly can gather, analyze, synthesise, fuse and update intelligencefrom all sources on all aspects of an enemy in real time and make it useful to decisionmakers. This rolling product would be returned to agencies and to an office of an UnderSecretary of Defense Intelligence, with three analytical-operational functions, “Warning,CI/Force Protection and War-fighting intelligence”. 15 At every level of command fromPentagon to platoon, decision makers would receive and discuss all relevant intelligencethrough “horizontal fusion”. 16 At first glance, such a body might solve the problemscaused by uncoordinated single source agencies, but not that of information overload inanalysis, unless one assumes that to centralise and automate and computerise informationmust transform its nature. Precisely that is the assumption. As its name indicates, the IDCis intended to unleash the power of information: to bring intelligence into the informationrevolution, and vice versa. Here, as often in the debates over the RMA andintelligence, vague language and jargon obscures a clash between ideas and agencies.Advocates of the RMA view the intelligence services which survived the cold war aslegacy forces, industrial age dinosaurs, muscle bound and clumsy, too focused ontechnique, security, secrecy and the source of their collection as against the material itprovides; too divided in acquiring their evidence and presenting their analyses; tooreluctant to disseminate their data; providing too much useless information; too little ableto answer key questions fast and accurately. In the cold war, American intelligencefocused on supporting millions of soldiers in a world wide competition against a peer,14 “Speech to the Heritage Foundation”, 13.5.03, by Arthur Cebrowski, Transformation Trends, 27.5.03,www.oft.osd.mil/15 Arthur Cebrowski, 17.6.03, “The Path Not Taken...yet”, www.oft.osd.mil/16 Rudi Williams, “’Horizontal Fusion’ Makes Troops Less Vulnerable, More Lethal”, American ForcesInformation Service, 26.9.03.
with the trump suit being the collection of data on strategic issues through technicalmeans; in the information age, the foci are terrorists or expeditionary forces. To meetthese needs, the revolutionaries want intelligence services to become nimble, to simplifytheir techniques and reduce their emphasis on them, to alter their priorities and their focuson one source; to cease being monopolists of knowledge and oracles of assessment; todistribute their material widely and freely, to fuse it constantly in a rolling fashion, tocooperate in assessment with each other and the military in ad hoc teams, and toemphasise broad strategic or political issues less and military operational ones more, toprovide less but better information. Advocates of the RMA see the attack on the twintowers as illustrating the flaws in American intelligence, but their criticisms are morefundamental. They want a revolution in military intelligence. These ideas have political consequences, intended even if unspoken. If an IDC iscreated, one analytical bureau, a military one, reflecting its aims and means, will handleall information from all sources, and dominate analysis in the intelligence community. IfONA is practiced, power in analysis will move from Washington to theatre commands.All this will revive demarcation disputes between collectors and customers over prioritiesand between the military and the Central Intelligence Agency ( CIA ) in analysis. Again,the problems in American intelligence can be solved by many means, not merely thoseproposed by advocates of the RMA. Intelligence services can adapt to the times; they doso all of the time. Without transforming or detracting from other work, refined “push”and “pull” techniques should let them flexibly and immediately meet many of the needsof each of five divisions in an expedition, a good thing in itself. However, such reforms (even more those in the revolutionary programme ) raise the dangers of the militarizationand the tacticization of intelligence. Junior commanders always want to controlintelligence, more than ever in an era of C4ISR and expeditionary forces; yet such stepsthreaten to erode the advantages of critical mass or centralisation. Again, the rise of precision weapons and ISR blurs the boundary between targetacquisition and intelligence. After Afghanistan, one intelligence officer noted “Asweapon sys become more “intel centric” the importance of ISR increases proportionately...Think of Intel as a modern gun director”, while Myers said bombs “can be used likebullets from a rifle, aimed precisely and individually”. After Iraq, Cebrowski held “’thereal fight is a close-in sensor fight’, with fused intelligence and surveillance products thatreduce the number of steps and time between identifying a target and attacking it”; later,he advocated “an Intel/Surveillance based force”. 17 The concept of C4ISR has the virtueof placing its components in a process, each affecting and affected by each other, but ithas eroded older boundaries of meaning. Somehow, in moving from C3I to C4ISR,“computers” have eaten qualities once assigned to “command” while “intelligence” hasdiminished, as an idea connoting “to think” slips into one meaning “to sense”. The same17 Briefing Paper by Joe Mazzafro, “Operation Enduring Freedom, Intelligence Lessons Learned, AnUnofficial Quick Look”, JWAD Mini-Symposium, 7.5.02, www.maxwell.af.mil/au/awc/awcgate/awc-lesn.;Jim Garamone, “Myers says Joint Capabilities, Transformation Key to 21st Century War”, American ForcesInformation Service, 5.2.02, www.defenselink.mil/news/Feb2002/n020520002_200202054 ; Dawn. S.Onley and Susan M. Menke, “DOD seeks to exploit intelligence”, Government Computer News, 23.6.03;XXX.
link between sensors and intelligence also characterised the heart of the older intelligencesystem, the collection of data on Soviet nuclear forces, but since the point was notimmediate action ( indeed, precisely the opposite, to avoid it! ) there was breathing spacefor thought. This is less so when the aim is to immediately kill a soldier. When considering C4ISR, it is tempting to focus on the aspects most easilychanged, machines, to assume improvements to them must raise the performance of thehuman ones, or the whole, and to believe solutions to one set of problems ( targetacquisition) will solve another ( net assessment) . In fact, one can improve everytechnological aspect of C4ISR without aiding any of the human ones, possibly evenharming their performance. The same action can help target acquisition and harm netassessment. These pressures bolster the tendency in American intelligence to focus ontechnology. Some revolutionaries hold that only nonhuman means can allow a C4I andNCW system to work. Contributors to the Air Force 2025 project predicted a C4ISRsystem with the self-awareness of a man, or a god: “a series of intelligent microprocessor“brains”...all-knowing, all-sensing”; “an intelligence architecture with human-likecharacteristics. It will simultaneously sense and evaluate the earth in much the same wayyou remain aware of your day-to-day surroundings””. One theorist concludes, “Futuregenerations may come to regard tactical warfare as properly the business of machines andnot appropriate for people at all”. 18 Against this pressure is the cry for “humint”, but thistakes many forms. Many call for a change in the culture of intelligence; others on theneed for human intelligence, while soldiers define “humint” vaguely, meaning everythingfrom cultural awareness or linguistic knowledge, to a focus on humans as sources ofinformation and for subversion, to paying some attention to humans, or to anything buttechnology. Beneath the debate on intelligence is an inchoate struggle between emphasison humans or machines. Many of these ideas about intelligence are naive or misguided. Their advocatescannot achieve exactly what they intend. Still, their actions will have consequences. Theymay cause a revolution, even if not the one they plan. The RMA is the greatest matteraffecting intelligence today, and one of the greatest it has affected. Ideas on these issues have affected American intelligence services, most notablythose most closely linked to the military. Since 1995 they have focused increasingly on servingC4ISR, especially by improving their databases, links with customers and reachback. TheDefence Intelligence Agency ( DIA) aims to provide “Fused, Tailored IntelligenceEssential to Battlefield Dominance” and “Dominant Battlespace Intelligence for theWarfighter”. Among the five “Core Competencies” in the National Security Agency ( NSA) ’s“National Cryptologic Strategy for the 21st Century” is Goal 2-Military Operations, Ensure Dominant Battlespace Knowledge Through Integration of Cryptology with Joint Operations...18 “Preface”, “2025 In-Time Information Integrations System ( I3S)”, “The Man in the Chair: Cornerstoneof Global Battlespace Dominance”, “Wisdom Warfare for 2025”, Air Force 2025, 1996,www.au.af.mil/au; Thomas K. Adams, “Future Warfare and the Decline of Human Decisionmaking”,Parameters, Winter 2001-02.
(2) Integration of cryptology with point operations to ensure dominant battlespace knowledge. NSA seeks to: (a ) Anticipate warfighter intelligence needs--on time, anywhere, at the lowest possible classification... (3) Integration of cryptologic support to enable policy makers to promote stability and thwart aggression. NSA seeks to... (b) Work with policy customers to improve interoperability and ensure that intelligence can be tailored to meet customer needs. (c) Expand “pull” dissemination capabilities to enable customers to initiate real time requests to improve crisis support. (d) Work with the IC ( Intelligence Community) to create interactive databases that will enable searches for information gathered by members of the IC.In Afghanistan during 2002 ( and no doubt, Iraq in 2003 ) NSA personnel ,“integrated with thecombatant commander staffs...ensured field commanders and others had access to NSAoperations and crisis action centers; developed a collection system that supports military forcesabroad” and coordinated reachback. On 17 October 2002, the NSA’s Director, Michael Hayden,said, “As we speak, NSA has over 700 people-not producing SIGINT- but sitting in ourcustomer’s spaces explaining and sharing SIGINT”. The National Imagery and Mapping Agency( NIMA), too, deployed “the Target Management System Network”, giving its “customers directaccess to targeting support and navigation data from the NIMA precise point database”. 19 ByJanuary 2003, NIMA and NSA exchanged personnel and combined imagery, geospatialand signals intelligence at the point of first production, before it was sent to consumers.20After Operation Iraqi Freedom, the director of the CIA praised “the seamlessness, fusion,speed and quality of what is being provided on the battlefield” and held this proved hisorganisation must transform like other intelligence services. 21 These agencies all aim todistribute normal intelligence better than the best performance ever reached before. Though theyare reforming rather than transforming, this pressure may reinforce the military’s role inintelligence and the latter’s tendency to focus on technology, technique and tactics, despite therhetoric about the need to develop human sources and cultural awareness. Meanwhile, the intelligence community began to enter the information age. By2001 the United States government had several web-based but enclosed intelligenceintranets, linked to the military’s “SIPRnet” ( Secret Internet Protocol Router Network), aself-contained internet gated from the conventional one. Intelligence and governmentagencies were joined to “Intelink Intranet”, which had further sub- sections—“IntelinkCommonwealth” ( between American, Australian, British and Canadian agencies),19 “Vector 21, Defense Intelligence Agency Strategic Plan, 1999-2000”;www.loyola.edu/dept/politics/milintel ; NCS-21 ( National Cryptological Strategy for the 21st Century),www.nsa.goc/programs/ncs21/index.; Director of Central Intelligence, The 2002 Annual Report ofthe United States Intelligence Community, 1.03,www.cia.gov/cia/publications/Ann_Rpt_2002/index.; “Statement for the Record by LieutenantGeneral Michael V. Hayden, USAF, Director, National Security Agency/Chief, Central SecurityService, Before the Joint Inquiry of the Senate Select Conmmittee on Intelligence and the HousePermanent Select Committee on Intelligence, 17 October 2002”, www.nsa.gov/releases/speeches .20 Dan Caterinicchia, “NIMA, NSA increasing colloboration”, Federal Computer Weekly, 30.1.03.21 Dawn S. Onley, “Success in Iraq due to better info sharing, Tenet says”, Government Computer News,11.6.03.
“Intelink-SCI” for the “top-secret, compartmented intelligence level”, “ Intelink-S”, a“SIPRnet at the secret level” for military commands, and connections between the mainintelligence agencies and their consumers, like the CIA’s “Intelink PolicyNet” and theDIA’s “Joint Intelligence Virtual Architecture”, a web-based interactive system whichjoined 5300 analysts worldwide for normal work and reachback. These nets weresupported by steadily improving collection, search, and analysis functions which, thehead of the Joint Military Intelligence College noted, must allow “mining of data not onlyof what the analyst knows is important but also of-while unthought-of by the analyst-what might be of importance”. 22 By 2003, the agencies were beginning to deploy an“Intelligence Community System for Information Security system”, ( ICSIS) with securegateways for the transfer of messages between networks of different securityclassifications. 23 The inability of the CIA, the NSA and the Federal Bureau ofInvestigation to coordinate their databases before the attack on the twin towers shows thelimits to this work, but one should not overgeneralize from that failure. Even on 10September 2001, intelligence databases on traditional military and diplomatic mattersprobably were linked fairly well. No doubt they have been to work rather better since. These steps were in pursuit of greater visions. The “Strategic Investment Plan forIntelligence Community Analysis” noted that by 2007, the agencies aimed to achieve a virtual work environment enabled by collaborative and analytic tools, and interoperable databases...the breaking down of barriers and the sharing of databases of critical and common concern...a virtual work environment that connects databases across the IC, enabled by collaborative tools, policies, and a security framework to allow analysts to share knowledge and expertise and link them to collectors, consumers, allies, and outside experts...Efforts and electronic tracking and production systems to capture and make available intelligence “products” that can be recovered and reused by customers and other analysts ( knowledge warehouses).The aim was to create the “secure and classified sub-set” of the infosphere, “theintelsphere, which is the virtual knowledge repository of authoritative intelligenceinformation, relevant reference material, and resources used to store, maintain, access andprotect this information”. The DIA had taken the lead “in developing the conceptsunderpinning knowledge management in order to provide full battlespace visualization towarfighters and military planners”, but all military intelligence providers “are automatingtheir request, tasking, and response systems-at both the front and back ends-to serve ascattered and diverse constituency”, and creating “an integrated electronic productionenvironment. The large organizations, for example, are making strides, albeit somewhatuneven, in tracking customer requests and in recording and capturing production flow, aneffort that will become increasingly critical if we are to develop common “knowledgewarehouses” that are easily accessible to our customers and to each other”.22 “Vector 21”, ibid;. Frederick Thomas Martin, Top Secret Intranet : How U.S. Intelligence BuiltIntelink, The World’s Largest, Most Secure Network, ( 1998, New York); speech by A.Denis Clift,President, Joint Military Intelligence College, at Yale University, 27.4.02, “From Semaphore to Predator,Intelligence in the Internet Era”, www.DIA.MIL/Public/Testimonies/statement06 ; “Joint IntelligenceVirtual Architecture JIVA” www.fas.org/irp/program/core/jiva.23 Dawn S. Onley, “Intelligence analysts strive to share data”, Government Computer News, 5.28.03.
To this system must be allied search techniques which would sidestep informationoverload, and “reveal connections, facilitate analytic insights and deductions and stream-line search by prioritizing information, automatically populating databases, andintegrating data”. The ultimate aims were -Dynamically integrating national intelligence analysis from multiple sources with the timely reporting of tactical sensors, platforms, and other battlefield information. -Providing customer, user, and producer interfaces so that organizations at all levels ( national-allied/coalition-theater-tactical ) have access to digital data that each can retrieve, manipulate. -Using advanced models, architectures, automated metrics/management tools and authoritative production templates within a collaborative environment to dynamically assign, prioritize, track, and measure the operations/intelligence infosphere content.By 2010, the intelligence community hopes to have “a dynamic knowledge base...fullyaccessible from anywhere at any time by authorized users... Knowledge base linkage tocollectors with information needs/gaps automatically identified”. 24 So too, the Army’sDeputy Chief of Staff for Intelligence held that in the “knowledge-centric” army, analytic operations will be executed collaboratively in a distributive environment with extensive use of virtual-teaming capabilities. Analysts at every level and in multiple locations will come together in virtual analytic teams to satisfy unit of action and employment intelligence requirements. Each analyst will have access to the entire body of knowledge on the subject at hand and will draw on interactive, integrated, interoperable databases to rapidly enable understanding. Communities of analytic interest will create and collapse around individual issues. The commander forward will be supported by the entire power of the formerly echeloned, hierarchical, analytic team. 25 No doubt, some of this is mere rhetoric, a political response to pressure, whilemuch that matters remains unsaid. Collectors and analysts fear the uncontrolleddistribution of their best material, or that transformation might degrade their work.According to an informed commentator, Bruce Berkowitz, in 2001-02 the CIA’sDirectorates of Intelligence ( DI) and Operations ( DO), had incompatible databases.Each DI analyst had one computer linked to their own ( poor) database, another to theinternet, neither to other official systems. “The CIA view is that there are risks toconnecting CIA systems even to classified systems elsewhere” . Merely to send intranetemail to intelligence officers outside the agency was hard. Few CIA computers werelinked to SIPRnet, though models which could receive but not send messages werequickly being introduced. The CIA disliked Intelink because it could not controldissemination of documents sent there. It did “post almost all of its products onCIASource, a website maintained on the Agency’s network that is linked to Intelink”, to24 ADCI/AP 2000-01, “Strategic Investment Plan for Intelligence CommunityAnalysis”, www.cia.gov/cia/publications/pub25 Lt. Gen. Robert W. Noonan, Jr., and Lt. Col. Brad T. Andrew, Retired, “Army Intelligence Provides theKnowledge Edge”, Army Magazine, April 2002.
which few outsiders had access ( this required approval both of a person and a computer).No one outside the agency had much electronic access to CIA material. In order to studyany topic, DI analysts had to search separate databases, on the DI system, Intelink, andthe internet, and often simply ignored the last two. “When it comes to IT, the CIA’sapproach is not “risk management” but “risk exclusion’”. All this had cultural causes andconsequences. Access by outsiders to CIA data threatened its hierarchical system ofassessment and quality control, while “by making technology a bogey-man rather thanan ally, the CIA is reinforcing the well-known tendency toward introversion among mostDI analysts”.’ 26 This critique was accurate for the time, but changes seem to haveoccurred since. By 2003, the CIA claimed to be creating secure but flexible inter-agencydatabases and intranets, including the ICSIS ( which had languished since 1998 ) and abrowser based system allowing document sharing and e-mail. 27 So too, in 1999, Hayden had two teams investigate the NSA. Both condemned itsculture and rejection of the internet. The internal team noted “we focus more on our own“tradecraft” than on our customers, partners, and stakeholders”. It advocated the NSA’s“transformation... from an industrial age monopoly to an information age organizationthat has entered the competitive market place”, embracing “the Internet as a force-multiplier... a means of creating numerous virtual centers of excellence with colleaguesaround the world”. The external team held that the NSA must be recentred around theinternet, and overcome its “ culture that discourages sending bad news up the chain ofcommand...a society where people were afraid to express their own thoughts”. “NSAgenerally talks like engineers. NSA will talk about the technic parameters of constructinga watch, describing gears and springs, when the customer simply wants to understand thatwe have just developed a better way to tell time. Even more important, NSA needs tolearn to communicate what the ability to tell time might mean to a customer”. 28 Advocates of the information revolution view the CIA and NSA as anal andossified, unable to see the need to provide fast, flexible, fused material. Such criticismshave force. Yet, Hayden noted, the NSA “is a very conservative, risk-averseorganization” because these characteristics, and “consistency and thoroughness andcare”, fit the cold war. 29 They also suit any intelligence agency which dislikes error orinsecurity. The rhetoric of transformation obscures the key issue—who gets what fromwhom? It is easy enough to produce good intelligence or coordinate analysts, or pushmaterial effectively in a crisis, or send fused information fast from Washington to atheatre command, or from there to an aircraft or unit commander. These are standardproblems with school solutions: reachback, tailoring, fusion, and pull techniques are oldhat, they do not require transformation. To give thousands of people access to the26 Bruce Berkowitz, “The DI and “IT”, Failing to Keep Up With the Information Revolution”, Studies inIntelligence, 47/1, 5.03, www.cia.gov/csi/studies/vol47/no1/article0727 Wilson P. Dizzard III, “White Houses promotes data sharing” and Dawn S. Onley, “Intelligence analystsstrive to share data”, Government Computer News, 28.6.02 and 5.28.03.28 External Team Report, A Management Review for the Director, NSA , 11.10.99; New EnterpriseTeam, ( NETeam) Recommendations, The Director’s Work Plan for Change, 1.10.99,www.nsa.gov/releases/reports.html29 Richard Lardner, “Leadership streamlined, chief of staff created, NSA Chief Pushes Ahead withOverhaul of Agency’s Culture, Operations”, Inside Defence Special Report, 16.10.00
databases of secret intelligence services, and to use them as they wish, however, isunprecedented, and an aim of the revolution. “Information typically has been retained inagencies’ channels until the product becomes intelligence”, said John Osterholtz of theDOD’s chief information office. ”We don’t want to hoard data until it is done. We wantto provide it as soon as possible”. 30 To do so, however, is to give intelligence servicescause to fear for their security and tradecraft and the integrity of their data. There areconcerns to be balanced. The question is, how ? Since 1995, the intelligence agencies have learned how better to push theirproduct to the military, which have honed their means to pull it. Reachback is a reality,which it will reshape. Already, it has magnified the power of intelligence. Colonels can tap datafrom the centre to solve problems in the field and guide immediate strikes; junior analysts athome can warn sergeants at the sharp end of danger or promise just ahead. Yet in all fairy tales,curses accompany blessings. Search engines augment analysis. They cannot replaceanalysts. Intelligence is a human process—it cannot simply be automated. Machinescannot command. C4ISR has solved only some problems of command and changed noneof its conditions. As ever, the issue is how much information a system contains, how fastand flexibly it circulates, and how well it is used. In communications, intelligence anddecision making, more or faster is not necessarily better; the value of multiplicationdepends on what is being multiplied; changes in quantity cause changes in quality,sometimes for the worse. If intelnets work as advertised, collectors can more easilydistribute their product and analysts find the material they want and correlations be made,expected or unexpected, and actions be aided--and everyone receive far more data thanever before, perhaps too much. So to justify their existence, intelink agencies will stock“knowledge factories” and an IDC with reports in mass, many trivial, some competitive,all well advertised; analysts will be swamped in sites, losing their way down hot-linkeddetours; analysis will be constipated by the quantity of information, conflicts ofinterpretation and of interest; need and politics will keep security restrictions in being,often blocking users precisely from the material they need or hiding all the best evidence.This situation calls for education in intelligence and its pathologies, and the creation of anew culture for assessment and use. If this aim is achieved, analysts and engines will justbe able to prevent the increased mass of detail from adding more friction to generaldecision making, while making great gains in two areas: when one knows what one wantsto know, the answers will come with unprecedented power and speed; and the chancesfor discovery by serendipity will rise. If this aim fails, more will mean worse. Again, by making intelligence more central than ever before, the United States has madeC4ISR the centre of gravity for its power and its greatest vulnerability. It also has increased theimportance and the difficulty of security. SIPRnet is the richest treasure ever for espionage, andintelink agencies its crown jewels, which will shine the brighter the better intelligence is fusedand distributed. In principle, the “intelsphere” is walled from enemies but accessible to friends,who communicate with freedom. If this firewall collapses, that web-based communication willbecome the equivalent of plain language traffic over wireless. If an enemy penetrates the“intelsphere”, it will have more chances than ever before to gather intelligence on you or to pirateyour intelligence and use it as its own, and an unprecedented ability to corrupt your data. Thatdanger, whether occurring in secret, in known form or even just as a suspicion, right or wrong , is30 Frank Tiboni, “Quantum Leap tests network warfare”. Federal Computer Weekly, 27.8.03
the achille’s heel of the “intelsphere”. If soldiers cannot trust the “intelsphere”, how will they acton it? One successful corruption of information, producing one failure in the field, might cripplethe machine or the trust on which it relies. A virus may be little more damaging than the fear ofone. Thus, as ever, security will trump flexibility. Databases are most easily defendedwhen they are accessible only to regulated computers. SIPRnet, however, is easilyaccessible—anyone capturing intact any one of thousands of vehicles in Operation IraqiFreedom in theory could reach any database linked to it. Penetration is common. In 2001,16,000 attempts were made to enter United States Navy computer networks, “of which400 gained entry, and 40 traveled the networks”. In 2002, 29 attacks on USAF networks“resulted in some compromise of information or denial of service”. In 2003, theAmerican Chief of Naval Operations, Admiral Vern Clark, said, “If you are a parenttoday, raising a youngster, one of the areas you ought to push them toward is networksecurity. This is going to be a job market that is crying out for people for the next 200years”. 31 This danger will limit the significance of intelligence easily reached throughSIPRnet, and force intelligence agencies to shelter their best material behind secure gatesor else on separate intranets. The “intelsphere” must stand apart from the “infosphere”,while still being linked to it through secure and flexible procedures, which allowinformation to be pushed from the top and pulled from below, and more coordinationbetween databases, and between analysts and users. These links may be more rich, thick,flexible and fast than ever before, or alternately below the standards of 1944 or even1918, depending on the relationship between the techniques of cyber attack and defence.At worst, C4ISR may follow the classic downward spiral of C3I over radio, where jamming andthe need for security sapped most of its flexibility and power. The mechanisation of intelligence and command also has transformed the securitydilemma. The new killer applications are spies to steal information and cyberwar to corruptdatabases. The key danger from hackers is less of an ULTRA than a nuclear strike on data; anagent in place, conversely, could betray one’s entire database of intelligence and command. In thecold war, sergeants turned spies pillaged storehouses of paper secrets—now, a Walker familycould loot “knowledge warehouses”, or corrupt them. Cyber defence must be geared to handleevery possible enemy everywhere all of the time, though for the United States mercenary hackersmay be the greatest problem for decades, until the rise of a peer competitor. Other states will haveto reckon with peers, or superiors. American authorities recognise these threats and act againstthem. Thus, in 2002, “a layered cyber-defense system” protected the Defense InformationInfrastructure, combining “local DoD intrusion detection systems” with an NSA controlled“computer network defense intrusion detection system... a network of sensors that arestrategically placed within the DoD infrastructure, providing analysts the capability to identifyanomalous cyber activities traversing the network”. The Pentagon hosts annual cyberwarcompetitions which, in 2003, included “a so-called rogue box in each network that the red teamcould use to simulate insider attacks”. 32 Though no American headquarters ever may feature31 Charles L. Munns, “Another View: Navy’s network services buy pays off”, Government ComputerNews, 3.7.03; David A. Fulghum and Douglas Barrie, “Cracks in the Net, Foes of the West Look forchinks in its technological armor”, Aviation News & Space Technology, 30.6.03; “Statement by JohnGilligan, Chief Information Officer, United States Air Force, Before the SubCommittee on Terrorism,Unconventional Threats and Capabilities House Armed Service Committee”, 3.4.03,www.house.gov/hasc/openingstatementsandpressreleases/10thcongress/02-04-03gil.32 DCI, “2002 Annual Report”, ibid; William Jackson, “Cyberdrill carries over to real war”,Government Computer News, 19.5.03.
signs reading “He who uses the computer is a traitor!”, this will happen to smaller states.The electron is a weapon. It can be used by you as well as against you. Compared to C4ISR and NCW, IO is a less novel and less problematical concept.It describes its subject better than any extant term, like “covert action”. IO embracesmany “disciplines”—deception, operational security, electronic warfare andpsychological operations, but also civil and public affairs, that is, public and pressrelations. Initially, the latter were added to IO to meet the Army’s distinct problems withpeacekeeping, but the relationship grew to include political warfare, both defensive andoffensive. During the campaign in Afghanistan, the Pentagon briefed journalists about thetechniques of Serb and Iraqi propaganda and press manipulation, or “enemy denial anddeception”. 33 This concern shaped its media policy during the 2003 war in Iraq, and alsoled to the short lived “Office of Strategic Influence” of 2002, a military organisation toshape international media coverage. Due to bad publicity, that office closed a day after itsexistence was announced, and no doubt reopened the next under a new title. ThatAmerican doctrine about IO fuses in one category matters once treated as “black” (psyops) and “white” ( public relations) and regards their combined practice as normal,presents problems for journalists, the public, and the military itself. With the significantexception of Computer Network Attack ( CNA), IO does not involve pouring old wineinto new bottles, merely placing new labels on old bottles. Functions which intelligenceofficers once might have conducted in a General Staff, perhaps with operations, securityand signals personnel in secondary roles, are now treated as a combat arm, controlled bythe senior Operations officer, with intelligence personnel first among equals of specialistelements. This rise of Operations and decline of Intelligence is marginal and reasonable;IO are operational matters, but in need of a close relationship with intelligence and otherelements. The basic doctrine for IO is sound, and close to the best practices of the bestpractitioners of two world wars. IO should be controlled by an officer directlyresponsible to a commander, guided by a small “cell” of specialists, able to provideexpertise and liaison, deliberately organised in an ad hoc manner, cut to fit the cloth; thevarious “disciplines” of IO should be “fused”; not merely coordinated, but combined.34 Thus, American doctrine on deception regards all aspects of intelligence as forcemultipliers, to be integrated into every aspect of planning and operations. It defines soundprinciples — “centralized control”, “security”; “timeliness” in planning and execution;“integration” of deceit with an operation; and, above all, “focus” and “objective”, aimingto influence the right decision makers and to affect their actions—to treat themanipulation of intelligence and ideas merely as means to an end. In order to achievethese ends, practitioners must understand their foe’s psychology, “possess fertileimaginations and the ability to be creative”; they must pass a story through many sourceswhich an adversary will find believable, ideally by reinforcing its expectations; and fuse33 “Background Briefing on Enemy Denial and Deception”, 24.10.01,www.defenselink.mil/news/Oct2001/t10242001_t1024dd.ht34 Joint Chiefs of Staff, Joint Pub 3-58, Joint Doctrine for Military Deception, 31.5.96 ( under revision asof time of writing, June 2003); Joint Pub 3-54, Joint Doctrine for Operations Security, 24.2.97; Joint Pub3-13, Joint Doctrine for Information Operations, JCS, 9.10.98; Joint Publication 2-01.3, 24.5.00, JointTactics, Techniques, and Procedures for Joint Intelligence Preparation of the Battlespace.
intelligence, psychological warfare and operations security with deception. This doctrineis powerful, but it has weaknesses which stem from the roots of its strength, the influenceof the British tradition of deception. The campaigns of 1943-44 which culminated inFORTITUDE stem from so many unique circumstances that they are a poor guide to theaverage. To treat them as normal is to assume deception is precise and predictable, thatone will have edges equivalent to ULTRA and the “double cross system”, while theenemy’s intelligence is castrated. These are tall assumptions. Again, “focus” and“objective” are fine principles: but in order to make key decision makers act as onewishes, one must know who they are, what they expect, how to reach them and how toknow whether one has succeeded. This is not easy. Deceivers wrestle with uncertaintiesand pull strings they hope are attached to levers in a complex system they do notunderstand. Deception rarely has just the effect one wants and nothing else. Theunintended cannot be avoided. American doctrine urges that this difficulty and others beresolved through risk assessment, but that is to mistake a condition for a problem. Reasonis good, war games are fun; when assessment concludes, risks remain.35 IO doctrine has been easier to write than to test. American experience with itscomponents since 1989 ranged in quality from poor ( Somalia) to decent but uninspired (Panama; the 1991 Gulf War; Kosovo). In 1998, one IO colonel with experience in Bosnia and atFort Leavenworth, Craig Jones, noted “much confusion remains--IO is still many different thingsto many different people”. It lacked measures of effectiveness. “ A commander has the right andthe responsibility to ask his IO staff officer this simple question: "How do we know this IO stuffis helping me achieve my overall objectives?" Between the conception of the idea and 2003, theUS military had no experience with IO in war, except in Kosovo, where Serbs matchedAmericans; its theory was drawn from history, where good examples did abound, some stillbetter than Operation Iraqi Freedom. The theory was good—the problem was praxis. In 2000 theIO Franchise, Battle Command Battle Lab, at Training and Doctrine Command, admitted theneed for basic studies on IO in war, including “good, reliable means” to assess its impact.“Intelligence doctrine addressing IO remains to be produced, and training remains concentratedon the traditional functions of locating and identifying opposing forces”, with “intelligenceproducts... designed to support force-on-force, kinetic, lethal engagements”. Again, in“numerous Army Warfighter Exercises... attrition-focused command and staff training exercises”,time was too short “to employ the less tangible aspects of IO in a manner that would influencethe operation”, and to understand their value and limits. A division or corps headquarters had justhours “between receipt of mission to course of action selection, allowing as much time aspossible for coordination, synchronization, and orders production”, and one at theatre level “outto 120 hours and beyond”. Deception and psyops, however, might take months to work. “Themost important aspect of IO at the tactical and operational levels is execution”, yet how far coulddivision or corps staffs “effectively integrate and execute all elements of IO into their decision-making processes, given the time constraints common to tactical operations?” If not, how shouldIO be organised? Historical experience, incidentally, including that from Iraq in 2003, suggestsdeception and psyops can work well even within a month from their start, while a theatre level35 John Ferris, “The Roots of FORTITUDE: The Evolution of British Deception in the Second World War”,in T.G. Mahnken (ed), The Paradox of Intelligence: Essays in Honour of Michael Handel ( Frank Cass ,London, 2003).
headquarters should handle these matters for subordinate commands. Again, how should olddisciplines like EW be adapted to fit new technology, IO and the information age? 36 Even experts were unsure how to apply IO. It was practiced only in exercises,peacekeeping, and Kosovo, all experiences with limits. Bosnia illuminated IO’s use inpeacekeeping but not in war, and perhaps provided some counterproductive lessons: adirecting committee of 20-plus members possibly is too bureaucratic for operations. In2000, reflecting on experience with IO in divisional work at the Exercise and Training IntegrationCenter, one IO Analyst, Roy Hollis, noted “All too often, IO is associated with rear area or forceprotection operations only... This is only a part of what IO is capable of doing, and we have tounlearn this”. Since “many staffers do not fully understand or appreciate the value of IO”,without effective “IO staff huddles”, “strong leadership” from the IO officer and a supportiveChief of Staff, “the staff will focus on what they already know and give minimal attention toInformation Operation requirements”. Personnel “lacked actual subject matter expertise in thevarious disciplines or elements, plus intelligence support, that make up Information Operations”,and spent too much time in too many meetings. 37 Amateurishness and bureaucracy are commonproblems in military intelligence; but for IO, a focus on form may turn revolution into a checklist.Until 2003, these problems characterised American efforts to apply their IO doctrine. Properly handled, IO are powerful tools, but they do not necessarily work as onehopes, and they can be used by one’s adversary. Defence matters as much as attack; itsimply is harder. Their power will be multiplied in an unpredictable way by the rise of anew discipline. Unclassified material rarely refers to CNA but the topic has not beenignored, simply treated with secrecy, just as armies did deception and signals intelligencebetween 1919-39. One USAF intelligence officer notes “offensive IO weapons...remainshrouded in limited-access programs” ; the Joint Chiefs of Staff’s doctrine on IOdiscusses CNA in a classified annexe; in 2000-01, the USAF sponsored research intospecialist “Cyber-Warfare Forces” , “potential targeting issues” and “how to mitigate orminimize collateral damage effects”, how CNA would affect “the full-spectrum ofInformation Attacks ” and create new “broadly defined multi-disciplinary activities, suchas: cyber-based deception, Electro-Magnetic Interference ( EMI), Web Security,Perception Management. How do we integrate/fuse input and provide a COA (Course ofAction)?”. 38 The Pentagon’s Command and Control Research Program describes CNAas “a rapidly evolving field of study with its own concepts and technology”. 39 Anyoneable to employ a hacker for love or money can hope to gain from CNA, while attacksomewhere is easier than defence everywhere. The entry costs are small, the potential36 Center for Army Lessons Learned ( CALL) 2 October 2000, Information Operations Franchise ,Research Project Proposal #s 4, 5.9, 10 and 12, www.call.army.mil/io/research37 CALL, “The Information Operations Process”; “Tactics, Techniques and Procedures forInformation Operations (IO). Information Operations, Observations, TTP, and Lessons Learned.www.call.army.mil/io/ll .38 Research Topics proposed by INNS, 25.7.00, “Information Operations ( IO ) 5.16”, Computer NetworkWarfare, “Information Operations ( IO ) 5.23, and 5.29 ; and Air Force Materiel Command, 8.28.01,“Effects Based Information Operations”, www.research.maxwell.af.mil/js_Database ; Colonel Carla D.Bass, “Building Castles on Sand, Underestimating the Tide of Information Operations, Aerospace PowerJournal, Summer 1999.39 “CCRP Initiatives”, Office of the Assistant Secretary of Defense Command, Control, Communicationsand Intelligence, 9.26.2001, www.dodccrp.org
payoff large, and the consequences uncertain. Sooner or later some state will let slip thebytes of cyberwar, with uncertain effect. CNA may revolutionize IO by incapacitatingcomputer systems, or replacing true data with false; or it may prove Y2K revisited. Afirst strike may be so advantageous that it creates an imperative to move first, adding anew twist to deterrence. Even when not used, CNA creates, one veteran noted, “built-inparanoia”, the need to fear that hostile states or non-state actors are attacking and to reacton that assumption to anything which looks like a threat40-- doubly so since CNA may beindistinguishable from accident, its authors undetectable, and inflict mass destruction (consider the consequences of wrecking the computers controlling air traffic control atHeathrow Airport, or of a nuclear power plant). So too, the nature of power in CNA isunknown: “How do you measure IO power?”, asks the USAF’s Institute for NationalSecurity Studies; “How would one calculate Correlation of Forces a la pastSoviet/Russian approaches?”; what are the “units of IW force” or their structure “e.g.squadrons of IW computers”. 41 IO is a known commodity; not so, C4ISR and NCW. How will reachback,intelnets, an IDC or “knowledge warehouses” affect the normal working of intelligence,and its use in crises or operations? Certainly, they will not end uncertainty, but just createa new kind— what Michael Handel called “Type B uncertainty”, the problem of decisionmaking in a context of too much and too constantly changing information. 42 Uncertaintyis not just about what is seen, but with how we see; not merely what we know, but howwe know that we know what we know; because of too few facts, and too many. It is acondition linked to problems. The problems can be eliminated, but attempts to end oneoften create another. A condition of uncertainty is that one can never solve all of itsproblems at any one time; merely choose which problems to avoid, or embrace; and theconditions must be endured. One can increase one’s certainty, and reduce that of anadversary or gain an advantage over it, and these gains may be great; but none of this iseasy to achieve. When facing a serious foe, uncertainty will remain sizable. Even againsta weak one, it can never vanish—chess players, knowing their foe’s dispositions, remainuncertain about its intentions and the clash of their own strategies. C4ISR and “DominantBattlespace Knowledge” ( DBK) will increase uncertainty precisely through the way theyreduce it; so too friction. In time of routine, they will provide more data than a generalneeds. In time of crisis they will produce less intelligence. How far will the ability tocollect and process information under routine circumstances affect ideas of whatintelligence can do when it matters? Will such a routine not merely hide pathologies andparadoxes and make them even more debilitating when they strike?—which will be whenit matters. What will a machine relying on the receipt of facts in hosts do if deprived ofthem? how will information junkies behave when thrown into cold turkey—just whenbattle starts? To date, the answers are unpleasant. The USN’s “Global 2000” war games testedthe application of NCW. It found both the power of C3I and its classic problems40 Dan Caternicchia, “DOD forms cyberattack task force”, Federal Compuer Weekly, 10.2.03.41 Research Topics proposed by INNS 25.7.00, IO 5.48, and IO 5.24, ibid.42 Ferris, J.R., and Michael Handel, "Clausewitz, Intelligence, Uncertainty and the Art of Command inModern War", Intelligence and National Security, vol. 10, no. 1, January 1995, pp. 1-58.
multiplied; with every member of the net able to post and edit notes, informationoverload paralysed command—officers had so much data that they could use little of it,bad coin drove out good. One witness questioned the validity of “visions of a command-and-control structure akin to the civilian internet... that the natural creativity, spontaneity,and adaptability of war fighters can be unleashed by freedom from constraint analogousto that of the civilian Internet in commercial settings”. 43 Experience in the Kosovocampaign led Air Commodore Stuart Peach to sombre conclusions: “the drive tostreamline procedures and handle ever more data has had an important side effect; airmenhave become driven by process not strategy”, “in reality, theory, doctrine and practicecollide with process. Airmen claim one thing ( centralized command and decentralizedexecution ) and in fact practice another ( centralized command and centralizedexecution)”; “refining the process of airspace control orders, air tasking orders and airtask messages became the performance criteria, rather than creative and bold operationalideas or campaign plans”. 44 According to a USAF officer, during this campaign theSupreme Allied Commander Europe, “had in his office a terminal that allowed him toview what Predator unmanned aerial vehicles in the air were seeing.” Once, whenWesley Clark viewed three vehicles he thought tanks, “he picked up a telephone, calledthe joint forces air component commander, and directed that those tanks be destroyed.With a single call, based on incomplete information, all the levels of war, from strategicto tactical, had been short-circuited”. 45 Similarly, during March 2002 in Afghanistan,officers in superior headquarters at home and abroad bombarded commanders withquestions and advice based on live pictures transmitted from Predators in flight. 46 A caseof friendly fire in that month showed that information overload, friction between layers ofcommand and inexperienced personnel, had swamped the USAF’s premier operational command,in western Asia. So much information was available that USAF squadrons could not circulatemuch material in ATOs to their pilots, while staff officers would not change their procedures,ensuring confusion between all layers of command. 47 The system processed and circulated farmore information faster than ever before, but in this high tempo environment, the need to spendjust thirty seconds in retrieving data could produce tragedy. It is so fast moving, fragile andcomplex that system errors are inevitable even without an enemy; the only questions are howoften, at what cost, and how much an enemy will multiply them. A key factor in any attempt to learn lessons is the difference between problemsand conditions. Problems can be solved, conditions must be endured. A fluid but43 Kenneth Watman, “Global 2000”, NWCR, Spring 2000. For a more optimistic reading of this exercise,and others, cf. Network Centric Warfare, Department of Defence Report to Congress, 27.7.01, pp. E-24, www.c3i.osd.mil/NCW/44 Air Commodore Stuart Peach, “The Airmen’s Dilemma: To Command or Control”, in Peter Gray, (ed)Air Power 21, Challenges for the New Century, Ministry of Defence, London: 2001, pp. 123-24, 141.An American Army observer, Timothy L. Thomas, offered similar views, “Kosovo and the Current Mythof Information Superiority”, Parameters, Spring 2000.45 Naval War College Review, Winter 2003, Major William A. Woodcock, “The Joint Forces AirConmmand Problem, Is Network-centric Warfare the Answer?”; p 46; the words are Woodcock’s, but hissource is Michael Short, the Joint Air Force commander in Kosovo.46 Anthony H. Cordesman, The Lessons of Afghanistan, Warfighting, Intelligence, ForceTransformation, Counterproliferation, and Arms Control, Center for Strategic and InternationalStudies, Washington, 12.8.02, pp 63-64.47 Verbatim Testimony of Colonel David. C. Nichols, and Colonel Laurence A. Stutzreim, Tarnack FarmsEnquiry. 1.03, www.barksdale.af.mil/tarnackfarms/rosenow.
hardened information and command system will not be easy to achieve. The aims mustbe to simplify the flood of data and direct it where needed, so avoiding the classicalproblem with satellite imagery, when one knew what to look for only after the start of thecrisis when that knowledge was needed. It will be hard to gain full access to data aboutknown unknowns and impossible about unknown unknowns. Nor can such systems workunless doctrine and training prepare people to use it. Still, one can reduce these newforms of uncertainty through old fashioned means. One must start by putting intelligencein its place. It does not make or execute decisions, people do, and more fundamentalissues—their education, intuition, doctrine, character, courage, openness of mind,wisdom, attitudes toward risk—determine how they understand and apply it. Knowledgeis only as useful as the action it inspires. Decision makers should listen to intelligence,and consider whether their perceptions are accurate, if they are pursuing the best meansto achieve their ends or noting all the salient points; yet they must remember thatintelligence cannot answer every question. They cannot wait for the last bit ofinformation to be received and for data processing to make their decisions. They mustknow when to act without intelligence or knowledge-- that is why they are leaders.Soldiers need to know well enough to act well enough when they must, and to understandwhen that moment is, no more, no less. The key questions are: what do you need toknow? when and how can you know that you know enough to act, or know that you knowall you can use? All shades of opinion recognise that C4ISR and DBK have magnifiedproblems like information overload, micromanagement and the fruitless search forcertainty, for which they share many solutions, such as changing the culture of command.Units must be able to operate in harmony without command, through some new versionof “marching to the sound of the guns”, or what the revolutionaries term “swarming”.Commanders must learn to act when they have a good enough picture of events evenwhen it is imperfect and new information is arriving, and to understand when they haveachieved that condition. Sometimes this process is called “to opticise”; Clausewitztermed a similar process the “imperative principle”. 48 When combined, these meanshave power and limits. They can solve many problems of command, perhaps most ofthem, but not all; and conditions will remain. C4ISR will be a function of a complexsystem manned by many people. It will suffer from all of the things natural to humansand complex systems, including uncertainty, friction, unachieved intentions, unintendedconsequences, unexpected failures and unplanned successes.NCW or C4ISR will not revolutionise ( nor even much change) events on the strategiclevel of war, or the strategic-diplomatic dimensions of peace, which are dominated byhuman rather than technological matters. Often they will affect such events counter-productively, by increasing confusion in and between levels of command. C4ISR andNCW sometimes will revolutionise tactics and operations where, all too often, friction atthe systematic level has reduced the value of intelligence; one actor had informationanother could have used but did not have in time to act, knowledge available in timecould not be used with effect; failures by any one cog prevented the whole machine fromworking well, or at all. In conventional war, NCW and C4ISR may ensure that every cogof the machine works well at the same time, reducing friction to the lowest level possible.All national intelligence assets will focus on giving every unit every chance to exploit48 ibid.
every fleeting opportunity; one’s forces will be used to asking for or receiving suchinformation and using it instantly, and well. In 1917, British signals intelligenceconstantly located U-Boats, prompting immediate air or surface strikes, which failedbecause units were slow and their ordnance weak. By 1943, intelligence on U-boats waslittle better but allied forces far more able to kill. In 1944-45, allied air forces, using thecab rank system, could strike any target reported immediately, if not accurately; in the2003 Iraq war, aircraft launched instant, precise and devastating strikes based oninformation acquired ten minutes earlier by headquarters 10000 miles away. C4ISR andNCW will raise the bar on the best use of intelligence, and the frequency of optimumuses, in conventional war; they will multiply any form of firepower relying on rapid,precise and long distance strikes, such as airpower. Little will change where equals engage, or the weaker side evades one’s strengthor strikes one’s C4ISR, or against guerrillas. If NCW fails in any instance on which it isrelied, disaster will be redoubled because of that fact; and fail NCW ultimately must. Ifsuccessful, it will force one’s adversaries to find solutions by evading your strength or bymaking you play to your weaknesses. It always is convenient when one’s enemy choosesto be foolish or weak, or foolish and weak, but sometimes it does not choose to be; andyou will be a fool to assume it must be so. A smart but weak foe may refuse any gamewhere you can apply your strengths, and make you play another one, such as terrorism. Atough and able foe might turn the characteristics of your game and machine into astrength of its own, by attacking any precondition for NCW and then by imposing itsrules on you. By doing what suits them in the context of our power, they will change theirstrengths and weaknesses—and yours too. The RMA has done many things, noteverything. It has multiplied American strengths but not reduced its weaknesses. It hasincreased the value of high technology and firepower in conventional war, but for littleelse; where these things matter, they do more than ever; where they do not, nothing haschanged. Iraq shows that the United States will aim to practice intelligence, commandand war at a higher level than ever achieved before. When it can play to its strengths, itwill succeed.John FerrisThe University of Calgary