SITUATION REPORT ON WARNING

• Collection: FOIA Collection
• Collection: Declassified Articles from Studies in Intelligence: The IC’s Journal for the Intelligence Professional
• Document Number (FOIA) /ESDN (CREST): 0000619200
• Document Page Count: 8
• Document Release Date: July 30, 2014
• Case Number: F-2013-02322
• Publication Date: December 1, 1983

Approved for Release: 2014/07/29 000619200TITLE: Situation Report on WarningAUTHOR: David Y. McManisVOLUME: 27 ISSUE: Winter YEAR: 1983Approved for Release: 2014/07/29 000619200(11roved for Release: 2014/07/29 000619200STUDIES IINTELLIGENCEA collection of articles on the historical, operational, doctrinal, and theoretical aspects of intelligence.All statements of fact, opinion or analysis expressed in Studies in Intelligence are those ofthe authors. They do not necessarily reflect official positions or views of the CentralIntelligence Agency or any other US Government entity, past or present. Nothing in thecontents should be construed as asserting or implying US Government endorsement of anarticle's factual statements and interpretations.Approved for Release: 2014/07/29 000619200Approved for Release: 2014/07/29 000619200'Data dumps vs. analysisSITUATION REPORT ON WARNINGDavid Y. McManisThe warning community, both governmental and academic, has spent agreat deal of time and effort looking at the warning -failures- of the past,starting with Pearl Harbor and carrying us through the invasion of Czechoslo-vakia and the fall of the Shah of Iran. While dwelling on past -failures" may-be academically instructive and even mildly rewarding, this is 1984, and not1941, and we as a community have evolved to a degree we probably neveranticipated.One way to understand how far we have come is to ga into Ivan's shoesjfor 'a look at the current US Warning System as he sees it. (I wish we had a bet-ter appreciation of how the Soviets really understand our system, just as I wishwe understood their warning system better. The technical components andwiring diagrams are relatively easy but it is much more difficult to understandthe cognitive warning process.)Ivan probably sees:a. an incredibly sophisticated total system replete with the latest incollection and information handling technology.b. continued rapid growth in collection capabilities with increasingemphasis on timeliness of processing.c. responsive and dynamic tasking of collection systems in responseto worldwide interests.d. a centralized command, control, and communications (C3)system which is without precedent or rival.e. a work force that is equally without precedence or parallel interms of its professional credentials and training.f. analytic tools which further amplify the capabilities of the workforce.g. reliable and competent allies with whom we are firmly linked interms of our C3 intelligence systems.h. an interaction with industry and academia on warning andintelligence problems in general, which forecasts continuedrapid improvements.Ivan might well conclude that the US system already has something like aStar-Wars level of technology: fast, centralized, cohesive, smoothly operating,and providing warnings and aiding executive response to events even beforethey reach crisis proportions.49Approved for Release: 2014/07/29 000619200Approved for Release: 2014/07/29 000619200WarningFundamentalFundamental Doctrinal ProblemsWe must keep firmly in mind that indeed our capabilities have developedenormously and we have much to be proud of. But so much for role playing,and now for the rude awakening. There are fundamental doctrinal problemswith which we confound our warning analysts.First, we have a very inadequate understanding of what "warning" reallyis and how it differs from conventional intelligence analysis. One means ofilluminating the distinction is to consider warning intelligence as having afocused goal which is the avoidance of harm; i.e., warning intelligence is thatwhich is produced early enough concerning a potentially threatening situation  to avert or ameliorate that forecast harm  (b)(1)(b)(3)(n)Seaond, there is a series of paradoxes which make warning reporting anextremely tedious business. Compartmentation is closely tied to the problemof distribution. A report which takes into full account our most sensitivesources of information can be issued to very few people. Conversely, to issue awarning report to a broad audience one must restrict the use of all-source in-formation, and thus an incomplete or inadequate warning may result. Timingand probability also work against one another. The nearer the possible event,the more likely we are to be able to accurately predict the probability of itsoccurrence. Conversely, to be able to avoid the harm implicit in the warningthe earliest possible lead time must be given, which again is related to a lowerlevel of confidence in the probability. Coordination should result in the mostcomplete presentation of relevant material, and normally the highest confi-dence level in the probability of occurrence. However, coordination tends tosuppress alternative analyses which may bode even more disastrous conse-quences for national security. Also, because our view of the future is perforcecloudy, it is hard to issue a coordinated warning sufficiently in advance of theevent to ensure that the harm can be ameliorated.Finally, our approach to the warning problem has been traditionally verynarrow, i.e., equated to indications and warning which has historically been amilitary game, with specific focus on the military threat from CommunistBloc countries. We are beginning to understand that these military indicatorscannot be viewed in isolation and indeed must be considered against thebackdrop of the world political and economic situation. This broadened visiondrives the complexity of the warning business to staggering dimensions.Technological ImplicationsLet us now address the implications of our rapidly evolving technologies.The collection and processing technology which has given us an unequalledcapability for reporting current events also has had a severe impact on ourconduct of longer term analysis. Today's analyst is no longer allowed theluxury of analysis; he has been forced to become a data processor and a50 SECRETApproved for Release: 2014/07/29 000619200Approved for Release: 2014/07/29 000619200Wprningcurrent events reporter. First, the quantity of data collected, even thoughthere is a significant amount of preprocessing, requires human review beforebecoming a part of a useful information resource. Second, the timeliness of thedata, and the partially understandable demands of the users, i.e., thepolicymakers and planners, force the analyst into a reactive mode where-'reporting what has just happened takes on greater importance than estimatingwhat may happen. The user does not consciously request this level of support,but through his actions and queries the analyst is forced into a defensivereporting role. In the "good old days" when analysts were working fromfragmentary information, and tasking for confirmation took weeks, if notmonths, the analyst was forced to rely on his studied knowledge of his target toprovide the best estimate of what might happen. Also because he had only lim-ited, and not very timely, insight into what had happened there was nopressure to beat the New York Times. Again, working with limited andmanageable inputs the analyst was able to build his own shoebox filesNrhichwere considered to be not only adequate but the only resPbpsive means to hislhistorical information needs. As a side excursion, the analyst also did not haveto worry about the security or integrity of his data base because it was underhis physical control, and even the smallest data bases were large enough to bevisible when removed from the premises.But, along came that villain technology to destroy the idyllic analyticexistence. Data began to pour in at an increasingly dizzying rate and analystsnow could quickly task another sensor for confirmation or amplification ratherthan make a judgment about the possible consequences of the event.Therefore, why speculate, interpolate, or extrapolate? The data base buildingproblem remained largely manual, at least to the extent of requiringsignificant human intervention, but because of the size of these basesadditional analysts were assigned to do nothing else. The target analyst lostcontrol and often even access. Undaunted, our clever analysts often recreatedtheir own pieces of the data base, which caused problems in redundancy andinconsistency. Yet we continued to improve our collection systems at anincreasing pace. There were constant collection inadequacies noted. Weneeded systems to see through walls and hear under water. The communica-tions spectrum grew and, as is always the case, the new spaces filled with data.We used these new communications capabilities to ensure that the dataarrived at our analyst's desk instantly.But alas, how was the technology helping the analyst, other than providinghim with more data to process and less access to information? The terminal ar-rived to replace the morning pile of messages and occasionally to replace hispen and paper. A few "automated" data bases became available, but thesewere often known for the difficulty of access and were rarely available from alocal terminal. To ensure access to more than one source of information theanalyst was required to learn the vagaries of each and every system to whichhe required access.What went wrong? Should we not have focused on improved collection?Were we not right in waiting for the perfect solution to the analyst'sproblems? The answers are not clear cut. But rather than dwell on what wentSECRET 51Approved for Release: 2014/07/29 000619200Approved for Release: 2014/07/29 000619200Warningwrong, let us consider what opportunities there are to meet the challenges ofthis era in information handling.Today, technology is not a problem; the rapid advancements in storage,processing, and display technologies virtually assure us of being able to meetthe analyst's requirements if they can be adequately defined and understood.The problem today is one of cognitive analysis, understanding how the analystdoes or should function and then providing the tools which will provide theanalyst access to the information and the necessary analytic routines.Will our current evolutionary approach do the job? I do not believe so-.Continuing to use the current analytic model will result in improvements inthe basic tools for processing, but will only make the analyst a better processorof data, not a better analyst of information. A new approach must beformulated which begins with the collection systems, viewing the strearrLofdata as something which wherever possible must be change() intn nc,41;1  information before being ? rovide11 1 1 . 1 .(b)(1)(b)(3)(n)The challenge, then, is not just to automate the way the analyst does hisbusiness today, but rather to analyze the basic functions of the analyst anddevelop new means of working with the information base necessary andpertinent to the task.TrainingIn scoping the responsibilities of the National Intelligence Officer forWarning, Mr. Casey and I agreed that it was the problems of process thatshould be addressed, and so I have undertaken initiatives in the followingareas.The training requirements inherent in the technological revolution lag inmany respects. Particularly at the middle levels of management and analysis,personnel are no longer familiar with the crises of the past and their inherentlessons to be learned. They are swept up in the daily routines of processing andcurrent intelligence reporting. Significant long-term analyses and projectionsare still being formulated but are often frustrated by outdated managerialunderstanding of the warning system or the warning process in general.Traditional forms of analysis cannot compete with the data flood. Our abidingconcern to be right has made us unduly wary of -false alarms- and the so-called -cry wolf- syndrome. The result is that good analysis is lost with the badApproved for Release: 2014/07/29 000619200Approved for Release: 2014/07/29 000619200warningand daring insights and projections never surface for analysts' fear of beingwrong, or they are too late to be of use except in hindsight.We are trying through dialogue with both the intelligence and operationalcommunities to develop a fuller and more subtle understanding of the use of.warning by those fulfilling different roles in the decision-making chain. Thisrequires new definitions and tools for warning based on a better understandingof how people react to threats.We have also established a program of community seminars to reviewwarning problems and possible solutions. The primary purpose is to spreadawareness of these problems among senior analysts and users and hope thatthey are able to convey the gospel within their agencies and departments. We,.are gaining valuable insights and notions of possible corrective actions fromthis sharing of ideas and experiences.We have established a Warning Training Working Group com-Ofisingrepresentatives from the Defense Intelligence College, Central IntelligencAgency, and the National Cryptological School to deal with the trainingrequirements for national warning. This group is currently developing awarning syllabus which will form the basis for new courses and be integratedinto existing courses. It is hoped that this syllabus can also form the outline fora "warning primer" which could be used for senior level policy officials whoare not, but must become familiar with the warning community, its proce-dures, and its products.We are beginning a program with the Army War College which we hopewill expand to the other service schools. Our intent is to foster a level of aca-demic research at the service schools which will expand our mutual under-standing of the warning problems.The art form for warning reporting needs a considerable amount ofattention. The only "official warning- vehicle is the alert memorandum,which suffers from g bad reputation and disuse. Our discussions with Mr.Casey and with community analysts and reporters lead us to the conclusionthat some form of warning memorandum may be required. It must have agreat flexibility to allow for any level of coordination and stimulation fromany level of the intelligence community. As with any warning vehicle, it musttake into full account the viability and importance of the "old boy network."A direct call from the DCI to the Chairman of the Joint Chiefs will alwayscarry more credibility than a formal memorandum to a community ofrecipients.Operations and Command Centers are a significant part of the warningapparatus. Most important is the recognition that as a network they have acapability that far transcends the sum of each of their individual capabilifi.cIn 1983 the chiefs of the Washington area centers met with me [(W(1 )  for a two-day conference. The dialogue wasextremely good and a number of actions have sprung from that meeting to in-clude upgrades to existing communications capabilities, the development oftailored training programs, and the development of a fiscal program fornetwork enhancement and training.SECRET 53Approved for Release: 2014/07/29 000619200Approved for Release: 2014/07/29 000619200?S-Ear WarningmrOther initiatives are in the talking stages to include looking at reportingmedia for warning; studying the applicability of disaster theory to warning;expansion of the Worldwide Indications Monitoring System to a nationalsystem including political and economic indicators; bringing our allies moreclosely into our warning system.For the future, many of the problems I have discussed will intensify. Butwe must avoid locking our understandings of warning into a straitjacket of thefamiliar. New understandings will be generated by new technology and newcrises. These represent opportunities for intelligent integration of automatedprocessing and manpower. Integration may require reorganization of the wayall of us do business. We must identify and foster genuine analysis, not just in-formation processing by our analysts. The automated tools will arrive, but theymust aid the analyst, not just complicate his job or narrow his vision.The Department of Defense has never lost its focus on the needwarning. Indeed, the DIA has been in the forefront of the development of in-dicator, methodologies, and now the message is spreading into other areas ofgovernment. But even the Department of Defense systems must continue toevolve. We must remember that the answers developed after Pearl Harborwere suitable for that era but are totally inadequate today. Tomorrow bringsonly more challenges and opportunities. Fortunately, the warning communityis stretching its limbs and is awake.This article is classifiedttettE-T.54Approved for Release: 2014/07/29 000619200