NEW ANALYST REPLIES

• Collection: FOIA Collection
• Collection: Declassified Articles from Studies in Intelligence: The IC’s Journal for the Intelligence Professional
• Document Number (FOIA) /ESDN (CREST): 0000620540
• Document Page Count: 10
• Document Release Date: July 30, 2014
• Case Number: F-2013-02322
• Publication Date: September 1, 1987

Approved for Release: 2014/07/29 000620540TITLE:A New AnalystRepliesAUTHOR:(b)(3)(c)VOLUME:31 ISSUE:FallYEAR:1987Approved for Release: 2014/07/29 000620540pproved for Release: 2014/07/29 000620540STUDIES 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 000620540Approved for Release: 2014/07/29 000620540Concentrate on substanceA NEW ANALYST REPLIES(b)(3)(c)CONF1bNTIALIn his  article -Managing/Training New Analysts- (Studies in Intelligence,Fall 1986),  (b)(3)(c)  makes three recommendations, which he states canfacilitate the new analyst's learning process. They are:? Communicate a sense of our mission and the difference betweenintelligence writing and academic writing.? Describe the process of intelligence analysis in a clear, cogent fashion.? Prepare the fledgling analysts for early failures and provide lots ofpositive reinforcement and reassurance.As a relatively-new analyst (three years with the Agency) with five years ofindustrial work experience, I would like to submit a few thoughts on thesubject.Academic Writing Versus Intelligence WritingI agree that substantial differences eicict hPtween academic and intelli-gence writing, and refer the reader to Mr.(b)(3)(c)  previous article for someexcellent examples of those differences. In particular, I believe there is muchtoo great an .emphasis on length rather than on substance in much academicwriting.In my own experience, I remember well (both in undergraduate andgraduate school) worrying if my research papers were substantial enough?innumber of pages, not only in substance. We students used to joke about ourreports being graded by the pound. In my experience, however, most professorsare themselves responsible for much of this verbosity. As an example, in thegraduate-level study program I am now enrolled in, each of the classes I havetaken so far has required submission of papers of specified length. In myopinion, emphasis on quantity rather than quality almost guarantees thatquantity at the exclusion of quality is what you will get.(In a somewhat ironic twist, I am now forced to read many of these verboseacademic-style research papers. Part of my analysis work depends on reportingby various components of NASA, which publishes its reports in the traditionalresearch paper style. It becomes very frustrating for me, now a consumer of thisreporting, to wade through the oceans of data and text to find the few drops ofconclusions I need.)Intelligence Writing: Not So UniqueIn contrasting intelligence and academic writing styles, however, Mr.Petersen implies that intelligence writing is unique. I believe that intelligenceCONFII.NTIAL 121Approved for Release: 2014/07/29 000620540Approved for Release: 2014/07/29 000620540CONQEJ4TIAL New Analystwriting is not a unique writing style, and that it does not require uniqueteaching techniques by intelligence managers. I submit that intelligencewriting is very similar to writing done for any large corporation or business.I offer my own work experience as an example. I was graduated fromcollege as a chemical engineer, and accepted a position in a factory of a largechemical company. I had no previous engineering work experience, and wasdefinitely "wet behind the ears." My job was to use my technical skills toimprove yields and throughput of our product. I quickly learned that some ofthe most important parts of my projects were the periodic reports sent to ourplant and division management. Let's see how these reports match up to theprevious paper's framework for intelligence writing.Focus on the FutureMy management didn't want to be bothered with details of previousprojects that had failed?it wanted to know what I was going to do now tomake a project run better. General business practices emphasize that sunk costs(past expenditures) are largely irrelevant; future costs and future rates of returnare the most important planning factors. Our management was planningcontinually 'for projected future business conditions, equipment purchases,maintenance needs, payroll requirements, and so forth. The past was onlyuseful for guiding us away from things that hadn't worked before, and givingus a basis for action in areas where we had been successful.Write for Generalists Who Are Grappling with Real ProblemsOne proverb often repeated around our office went something like this:"Reports should be short enough and clearly written enough that they caneasily be read and understood in the length of time of the average trip to thejohn" (roughly five minutes). In addition, we made sure that our conclusions,which always went on the first page, contained all of the important points,because we knew most managers wouldn't get any further than that first page.Industrial managers are just as busy as CIA managers, and can't afford the timeto dig through masses of data to find a few pieces of useful information.It was even more important for me to realize, as an engineer writingtechnical reports, that many managers in our factory were not technicallytrained. They were not interested in minute technological details of projects,but, on the other hand, they were responsible for making sound businessdecisions about those projects. Thus, these managers wanted reports thatincluded only essential details, written in a way they could understand without(God forbid) getting the idea that some technical smart aleck was trying to "talkdown" to them.Deal with Essentials Only, and Make Meaningful CharacterizationsThe temptation to become the "all-knowing expert" of whatever projectyou are working on is the same in industry and in the intelligence business. Thesame time and resource constraints also exist in both worlds. In one of my firstprojects, I was to increase throughput rates on a pigment grinding machine. Iinitially collected reams of data on every aspect of that machine. It became122 Co."'NFI?lag_t_ITIALApproved for Release: 2014/07/29 000620540Approved for Release: 2014/07/29 000620540New AnalystCC0111 SWIALapparent, however, that there were actually only a few key data points toconsider. When I reported results to my management, the necessary detailsbecame even fewer. I learned to recognize what was truly important in eachcase, and to spend only the time that was necessary to do the job right at theappropriate level of detail.Our project reports always started with what we believed was essential formanagement to know. If a manager wanted to know more, he/she could readthe supporting details. If he/she wanted in-depth information about theproject, he/she could either call me up and ask about it (this very rarelyhappened) or request a briefing on the subject (even rarer).Begin with Conclusions and Then Explore Their ImplicationsUsing the above pigment grinding machine example again, in the long run,the increased throughput rate of that one pigment grinding machine was onlya small part of my project. Our company had many factories, all of which wereevaluating new ways of grinding pigment. It was my responsibility to compareour pigment grinding machine's new operating procedures with the otherfactories' current methods of operation, factoring in changes in equipmentusage rates, safety implications, long-term equipment reliability, and so forth.The implications of what the machine would do for our overall business weremuch more important than just the operation of the machine itself.Differences and SimilaritiesGiven, then, that industrial writing is similar to intelligence writing, whatcan be learned from an examination of a typical company's approach to theteaching of successful writing techniques? Analysis of my own experience leadsme to conclude that there is one important difference between industry andintelligence that leads to a lot of analyst and manager heartburn, and oneimportant similarity, which could lead to more effective writing by analysts.The Difference: EditingOne of the reasons managers (and analysts) come to believe that intelli-gence writing is a unique style is the tremendous edifice of editing that has beenbuilt up around it. It is my conviction that we have developed a structure thatis inefficient, and stifling to timely, accurate, and creative writing.This conviction became stronger as I became involved with the coordina-tion of my first paper (a joint effort with two other offices). The overwhelmingmajority of the drafts were returned with few or no comments of a substantivenature. Almost all of the drafts, however, were returned with the typographicalerrors circled and disagreements about grammatical style (which or who, fullybelieved or believed fully, and so forth) pointed out. What a frustratingexperience. I had expected questions and comments about substantive issues.What I received were everyone's efforts to be grammarians. From conversa-tions with fellow analysts, I have found that my experience was fairly typical.I'm not sure how this practice of misguided editing came about. I imaginebranch chiefs must feel a tremendous amount of pressure to make sure theCONFIbENTIAL 123Approved for Release: 2014/07/29 000620540Approved for Release: 2014/07/29 000620540CONFI NTIALNew Analystfinished intelligence writing that leaves their branches will not only betechnically correct, but grammatically elegant. I certainly do not condonegrammatical gaffes or obvious wrong writing, but I do feel that attention ismuch too heavily slanted toward the framework, and not on the picture theanalyst is trying to paint.Needless editing can also lead to adoption of particularly dangerousattitudes by analysts. If an analyst knows that, no matter how much time andeffort he or she puts into the construction of an article of finished intelligence,it will be ripped apart because of differences in personal grammatical styles, thetendency will be for the analyst to not want to try so hard next time. Why sweatthe details if someone else (or, perhaps more typically, many someone elses) isgoing to rewrite the piece anyway?- How Long Should It Take To Publish?Unnecessary editing can also significantly detract from the timeliness (andthus the value) of our finished intelligence. One analyst I know decided to keeptrack of the time his intelligence assessment took to get through the reviewprocess. Of the 10 months it took to publish the paper, approximately sevenmonths were spent in review by various managers. In my view, that is too long.If a mathematical equation were developed to quantify the value offinished intelligence reporting, the time between when the analysis is com-pleted and when the finished piece "hits the streets- certainly would be asignificant variable. This isn't news to any of us. Unfortunately, our reviewprocess often doesn't seem to reflect the importance of timely finishedreporting. In our office, we even have an alternative to the regular paper reviewprocess, the so-called fast-track or streamlined paper review. This process isused for papers deemed to be particularly important or perishable. Isn't it sadthat we have to rely on special methods to ensure that analysis gets out in atimely fashion?So How Did You Do It in Industry?In my industrial job, my only editor was my immediate supervisor, a Ph.D.chemical engineer, who had engineering projects of his own as well as havingsupervisory responsibilities for nine engineers. He was too busy to worry aboutlooking up whether an engineer should have used "which- or "that-. He didensure that what went out of our group was technically accurate, however, bykeeping up to speed on the various projects on which we were working.How do you keep track of the multiple projects of nine engineers? Youdon't. That is, you don't keep track of the minutiae involved in every particularproject. It was our mutual responsibility, however, to keep each other abreastof significant developments (or lack or developments) in given projects.Moreover, my supervisor's level of involvement varied as my experiencelevel increased. At first, we worked very closely together. This was beneficialfor both of us. He was able to "show me the ropes- in a short period of time,and I wasn't left alone to figure everything out by myself. As my experiencegrew, so did his confidence in my abilities and judgment. I was left increasinglyon my own to decide the technical merits of my projects.124 CO1WJDENTIALApproved for Release: 2014/07/29 000620540Approved for Release: 2014/07/29 000620540New Analyst 0 ENTIALMy supervisor took the same approach toward my written reports as hedid to my project engineering work. At the beginning, he spent sufficient timewith me to make me comfortable with my company's report writing style. Hewent over the first few reports I wrote, with the attitude of making sure mylogic fit in with the desired format. After the first few reports, I was prettymuch on my own. My supervisor still saw my reports before they weredisseminated, but if they were accurate, he didn't make changes.What is my point? Simply that it is important not to let the process takeprecedence over the product. The purist may argue, correctly, that perhaps myreports were not as elegantly composed as they could have been. But I did putout reports that were timely, as well as technically accurate. They were goodenough. As time went on, I know my writing improved. The important thingwas, however, that I improved as I produced, as opposed to improving insteadof producing.One more thing. It was recognized by our plant management that mymanager was not responsible for my writing style. If the report was factuallycorrect, my manager had done his job. If the writing was not elegant, that wasmy fault, not my manager's.A Possible SolutionIn my opinion, employing the services of the central editorial staff withinthe DI earlier in the production process could go a long way toward solving theproblem of turning out timely, accurate reports. This staff in the Office ofCurrent Production and Analytical Support consists of professionally trainededitors, and is responsible for ensuring that all papers are published in the -DIstyle."The writing process I propose would go as follows. The analyst worksclosely with his/her branch chief and with substantive authorities for coordi-nation, in the preparation of the first draft. Close, frequent contact betweenanalyst and branch chief ensures as much as possible that everything stays ontrack and on focus. In this phase, the steps Mr.(b)(3)(C)  outlined in his-Framework for Analysis- can be particularly helpful.When the first draft is completed, the branch chief and the analyst satisfythemselves that the facts and the logic of the paper are correct, and that theyare getting their points across clearly. A certain amount of editing is necessaryin this stage, but it should be limited to that needed to ensure that the reportis accurate and understandable.Once the branch chief and the analyst are satisfied with the first draft, theysend it to the central editorial staff. This staff ensures that the paper reads -inthe DI style- and is understandable to the intended audience. The staff is infrequent contact with the analyst, so that questions about what the analystmeant by a particular statement, or how the analyst wants the graphics to look,can be resolved quickly.When the central editorial staff finishes with the paper, it announces thata paper on subject X has been completed, and anyone with knowledge in thisarea who would like to see the paper for comments can go to the centralCONKNTIAL 125Approved for Release: 2014/07/29 000620540Approved for Release: 2014/07/29 000620540CO>115EQTIAL New Analysteditorial staff office and do so. (Note: This is purely a precaution. Those withknowledge in the subject area were consulted in preparation of the draft. Theannouncement is just a way to ensure that all legitimate concerns are addressed,and that no one is trying to sneak a paper out without proper review.) The staffand the analyst who wrote the paper handle substantive comments, getting intouch with the commenting analyst as necessary. Stylistic comments arehandled between the staff and the commenting analyst. Once the commentsare resolved, the paper is published.This system would accomplish several tasks. First, it would remove thepressure on branch chiefs and analysts to overedit their work. Many of usagonize over our papers too much, with the result that we lose much of thetimeliness of the information. In our effort to make things look good, wesometimes forget that our product is not reports, it is information?informationthat quickly loses its value. A central staff handling hundreds of papers a yearwill be farther ahead on the learning curve than a single analyst, who publisheson the order of one paper per year, could ever hope to be. The timeliness andquality of our publications should both rise as a reNsult.Secondly, having all DI papers go through this central staff earlier wouldensure that all papers would be written in the same DI style. As even I, with myrelatively short tenure here, have learned, desired writing styles change overtime. It would be much faster, easier, and, I submit, cheaper, to effect thesechanges in all their subtle nuances with a small staff of professional writers thanwith hundreds of analysts.A third benefit would be that this central editorial staff would allow theanalysts more time to do analysis, at no expense to the Agency in terms of lostproduction of finished intelligence. The present system can require severalmonths of the analyst's time in poring over suggested changes, editorialcomments, draft revisions, and so forth?time that I believe could be betterspent in analyzing new problems. We could let the full-time writers polish thewriting, and let the analysts analyze.The Similarity: Learning by DoingI believe the best way to learn appropriate writing techniques is on the job.I learned to write in my company's particular nuance of the general industrialstyle by writing real reports, not by completing -make believe- assignments tohand in to my boss. I don't think many companies have the time to go throughthose kinds of exercises.I believe the same framework for constructive education about effectivewriting from one's own supervisor exists within our Agency. Most good branchchiefs are good writers (otherwise, they wouldn't have made it to branch chief).I believe the most effective way to train a new analyst to write correctly is tohave his/her own branch chief give a brief summary about why and how wewrite. The steps mentioned above and in the earlier article are an excellent wayto do this.Next, I would suggest that the new analyst begin writing short articles ofcurrent intelligence for some of our many publications. In my view, our126 CONNQENTIALApproved for Release: 2014/07/29 000620540Approved for Release: 2014/07/29 000620540New Analyst CO IDENTIALAgency is well set up for this type of writing instruction. In the Office ofScientific and Weapons Research, we have a daily publication for disseminat-ing current intelligence (the Science and Weapons Daily Review). This is agood forum for writing by new analysts. One of the most important things formanagers to remember is not to let the analyst get too hung up on the format.A review of other similar pieces should quickly give the analyst the desiredformat without raising excessive anxiety about getting the line length right, andso forth.Some may take exception to my position that the branch chief shouldteach writing. I feel very strongly that this is the most effective method. Ifbranch chiefs can't perform this function, they should be in a different position.Help for Branch ChiefsAs I mentioned above, I don't feel managers should be responsible foranalysts' writing styles. What, then, should be done about analysts who don'tseem to have the knaek for reporting intelligence in a way that is interesting forthe reader? Are these analysts doomed to poor performance reviews?No. But there's more than one way to skin a cat, and just because we're inthe intelligence business doesn't mean we can't go outside the Agency to get thehelp we need. This is the age of consultants, and one area where I feelconsultants can be particularly effective is in giving seminars on effectivewriting techniques. The company I worked for had regularly scheduled writingseminars, in which writing experts would offer their insight into what makesone style of writing interesting, another boring. OSWR management sponsoredsuch seminars in April 1986 and December 1987. I believe they were very wellreceived.New Analysts Do Not All Have the Same NeedsMrPD)(3)(C) ,lso addressed, albeit briefly, the need for mentoring, or, asit was called, hand-holding. I feel this is particularly important for new analystsof any age or experience level. I would make a careful distinction between thekinds of mentoring necessary, however. A new analyst straight out of collegemay need some initial assistance in adjusting to the work environment. Collegeis different from the workplace. There are different rules of etiquette to beobserved, and different expectations.The analyst who has had previous work experience, however, probablyalready has a handle on appropriate behavior patterns. These analysts needsomething closer to a quick walk-through of the -way things are done aroundhere,- an idea of performance expectations, and access to a good resourceperson who knows the nuts and bolts of the Agency's system of management.Most new analysts are, I would submit, eager to please, and only wish to bepointed in the right direction.Preparing the Fledgling Analyst for Failure?The third major point of the previous article was to prepare the fledglinganalysts for early failures and provide lots of positive reinforcement andCON NTIAL 127Approved for Release: 2014/07/29 000620540Approved for Release: 2014/07/29 000620540CONt&NTIAL New Analystreassurance. The author stated that "Each analyst must learn the job as eachbefore him has, essentially by trying, falling short, and trying again."I feel very strongly that "to prepare the fledgling analysts for earlyfailures" while providing "lots of positive reinforcement and reassurance" isexactly the wrong way to approach the training of new analysts. This approachis analogous to a trainer teaching a young pugilist to box by putting him in thering against the heavyweight champion. It doesn't matter how much encour-agement the trainer gives his pupil, the pupil is going to get pounded into thecanvas. When the cards are so stacked against you that you have no chance ofsucceeding, serious frustration is bound to set in. Systems that are set up forpeople to learn by failure and frustration are systems-that will lead to morefailure and frustration.My suggestion? Prepare fledgling analysts for early success. The earlierarticle contended that "In many cases, you will be the first person ever to tell[the new analysts] their work does not measure up." I disagree. Anyone who hasmade the transition from high school to college has had to adjust to differentexpectations. This Agency has a reputation for hiring intelligent people (no punintended), but I haven't seen anyone walking on water around here yet. We allhad to make some initial adjustments.Unforhinately, preparing new analysts for success requires more fore-thought and preparation by managers than the "learn by failure- mode.Preparing for success also requires more "up-front" training, so that, forexample, the branch chief knows that the analyst has the literary tools requiredto write a major paper by the time he has finished his research.We know, however, that the literary tools are only a part of what we needin order to succeed. I would not presume to speak as an expert on .the thequality of the Agency's management-analyst relationships. I have had someexperience as a foreman for production and shipping crews in my previousindustrial employment, so I have had the opportunity to learn a few thingsabout relationships with superiors, peers, and subordinates. One of the mostimportant lessons I learned was that if I wanted to get good performance frommy crew, I had to believe that they wanted to do a good job, that they werecapable of doing a good job, and that they would do a good job. Shortcomingsof my crew's performance in any of these areas meant that I had to reevaluatemy own performance to determine how I was keeping my employees fromdoing their best work. Employees tend naturally to rise (or sink) to your levelof confidence in them and your respect for their abilities. Statements such aswere made in the previous article about new analysts "driving you insane," orpunching this guy" for his incompetence, or having to "hammer" missionsand essences into the analyst's head,, have no place in management?even injest.Many authors have written books on effective management techniques?there are enough of them that one could certainly spend one's life just readingabout management and never doing it. I see no need in adding to the clutter.I would, however, in closing, like to pass on for consideration the most valuablepiece of management advice I have yet received. It goes like this: Do untoothers as you would have them do unto you.This article is classified CONFIIYNTIAL.128 CONF1.ENTIALApproved for Release: 2014/07/29 000620540