Ul UVlDUl l1 000lDWVU UlD V 15lD 15 W15 l 15 f $UllB W l1WWl I m uml1ml J 'il J jl 1J ill J 1 4 c P L 86-36 DATA SECURITY AND HUMAN ERROR 86-36 I I l zz 7 CONTINUING PROFESSIONALIZATION oooo o METHOD OF MEASURING NEGATIVE INTELLIGENCE oo 9 CAN'T TELL WHEAT FROM CHAFF oo oooo ooooooo J Gurin oooo oo ooo 11 REDUCTION RATIOS IN MICROGRAPHICS oooooooo Don Snow oooooo oooooo 12 BACK TO SQUARE ONE oooooooooooooooo oooo o 13 FEEDING THE GERMANS MISINFORMATION oooo ooooo P W Filby o oo oooooo 14 AND-A YOU BETtA HAVE MOTI-VAYSHl ooooooooooo t 15 NEWS OF COMMUNICATIONS ANALYSIS ASSOCIATION W E S ooo o o o o o 16 MORE B S BEFORE SPELLMAN oooo o 1 17 LETTERS TO EDITOR ooo oo oooooo o o o o oo Pattie Filby 19 NSA-CROSTIC NO 18 oooooooooooooooooo o o oo David H Wi11iams o 20 I 2 THIS 90Cl-JMEHT COHTMHS g09IllW0R9 MA ERIA 'n II 91RN8A 'QIIQ88 N8A 'Q88M 1 0iL hem 8B8 He 118111 eaRP I gee 'i' t Ipo Netilielltie Is the Qpjli_ter Declassified and Approved for Release by NSA on '10-'1 2- 20'1 2 pursuant to E O '135 26 vl DR Case # 54778 DOClD 4009815 Publi hed 'f6P SECRET Monthly by PI Techniques and Standards for the Personnel of Operations OCTOBER 1978 VOL o V No 10 PUBLISHER WILLIAM LUTWINIAK BOARD OF EDITORS Arthur J Salemme 5236s Editor in Chief Collection o oo 1 -- Cryptanalysis 1 Cryptolinguist ics o o ooooo 1 ---lIC855SS Language o o o o oo oo o 1 86-36 It4902s 1 5 1 8i lS 1 ' 5804S Mathematics o o o ooo o oo '- I I ' 7 8-5-1$S Machine Support ' o Special Research o o oo ooooo Vera R Filby 7119s Traffic Analysis o o oo Don Taurone 3S73 Production Manager oo o o o o o Harry Goff 5236s For individual subscriptions send name and organizational designator to CRYPTOLOG PI T9P SHeRi' DOCID 4009815 UNCLASSIFIED lJ RTfI 1 EE UH TYfl 1 ERR R w 1321 f P L I IIIUltratlon by tho author eople are not perfect they make mistakes In the computer world they make mistakes in operating machines entering data writing programs and designing systems In this article I will discuss some of the ways in which people's mistakes -- most of them caused by simple carelessness -- can cause major security violations to data in a computer-system environment The term data security refers to protection of data against persons against modifications or against destruction An understanding of the number of personnel and other resources having access to information is vital to estimating exposure probability You must also understand how personnel can access information From the security point of view a prime concern related to the data-preparation and data-control department is to protect the large quantities of data they handle from destruction and tampering of any kind Not only is there concern with the security of the data but also with the validity of the data Harm can be inflicted in several ways such as destroying or modifying other users' data reading or copying another user's data without knowing it and or degrading the service another user gets e g using up all the disk space or getting more than a fair share of the processing time An extreme case is an accident which crashes the system This might be considered the ultimate degradation P The security of information inside a computer system has been dealt with by several authors who have considered the problem from a number of different points of view ranging from privacy to government security All those authors stress the importance of peopZe in maintaining proper security Certain people will have access to the system by virtue of their position These include the computer operator programmers maintenance engineers users and certain management personnel Normally these people will be able to read modify or copy files and programs But all of them being peopZe can be careless and commit serious errors Let's look at some of them Keypunch Errors Keypunch errors can cause the loss of single records and modification of records To even the most trained eye incorrectly punched cards look the same as punched cards containing legitimate transactions But an error in keypunching can lead to destruction of data in a file Keypunch errors can occur when a programmar rushes to get one last run in before the end of the day In hurrying he dupes a card incorrectly throws the correct card 'away and adds the incorrect card to his deck Fig 1 -- c J o Fig 1 Other keypunch errors occur when the computer operator is tasked to repunch cards that were destroyed by the card reader Fig 2 Programmers and operators should take great care in examining the newly punched cards before entering them into the computer system TerminaZ Operator Input EPrors A user at a terminal accessing a file can cause loss of single records and modification of records Data preparation errors can cause a loss of data or permit erroneous data to enter the data file When a file is initially created the user of the file has unlimited access to perform any of a number of operations on the file These operations might include for example October 78 CRYPTOLOG Page 1 UNCLASSIFIED 86-36 DOCID 4009815 - --- ------------- UNCLASSIFIED 'J t - 'ffil JIl' t - ' Data can be destroyed in a file and replaced with inaccurate results leaving the user with a situation where it may take considerable time to realize that he no longer has the correct data The user is responsible for his share of the security problem and should restrict his terminal and batch processing of files to authorized and competent personnel only Program Errors Without a doubt the most common embarrassing and expensive software loss is program error A not fully debugged program coexisting with other programs might as well be regarded as having been written by a malicious enemy -- even if all the programs have the same author rogram bugs originate because of er rors in system design errors in logic errors in coding errors in problem definition or sometimes just incomplete definition Program errors or bugs often do not show up until some rare combination of circumstances reveals them They are so common that there is a regular mythology of program error often attributed to the computer instead of the program In some systems a few residual bugs do not matter too much in others they can be catastrophic Fig 2 Programs should be designed so that a user cannot alter the program If programs are permitted to be altered shared programs could be reading the file writing emptying renUmberin a constant state of flux and this would ing truncating deal locating unused disk lead to inconsistent results space and renaming files Fig 3 The majority of installations still have a real security gap when it comes to preventing insertion of unauthorized routines into programs ctf' 1 DOIIoI The Wa Street Journa reported the Borden Co pallY had announced a $2 8 million deficiency resulting from what appeared to be an error in switching part of the company accounting system over to computerization two years ago Dennis -'1 Van Tas sel Computer Seeurity Management p 54 At Cape Kennedy a space launching failed because of a program error The computer sy abol equivalent of a comma was inadvertently left out of the program The omission caused the rocket to go far off course and it had to be destroyed Ibid p 55 - t -ue While the impact of these two program errors is larger than most they illustrate the possible magnitude of the problem The prime answer to the problems mentioned is very thorough program testing Operator Errors Inexp rienced computer operators often ignore console error messages which can cause loss of single records modification of records and or loss of an entire file Fig 4 October 78 CRYPTOLOG Page 2 UNCLASSIFIED DOCID 4009815 UNCLASSIFIED An operator may accidentally process updates to a file twice thus erroneously updating files Throwing away the latest listing of an update and giving the user programmer an earlier listing can cause the user and or programmer to make erroneous changes to a file The following is an example of an operator error and what kind of problem it caused An operator mounted the wrong day's tapes The pro ram recognized the error and printed out a message saying so but the operator ignored the message and pushed the restart button This error was not discovered until month-end processing was completed after which much work had to be rerun More care should be taken by the computer operator to assure the accuracy and safety of files kept in the computer environment Programmer E1'1'ors Fig 4 The mounting of wrong tapes and or disk packs for updating can cause a loss of entire files and modification of records Physical damage to tapes or disk packs such as dropping them on the floor can cause loss of entire files or loss of single records Fig 5 Fig 5 Operators should be encouraged to strive for care rather than speed in handling tapes and or disk packs A programmer updating the wrong version of a program can cause a loss of an entire file and or modification of records A programmer who has access to the computer operations area can run his own jobs and make the same mistake an operator makes such as mounting wrong tapes ignoring console error messages and or carelessly dropping tapes or disk packs The following is an example of a programmer error and the damage it caused In August 1971 the French satellite Eole was launched by SA as part of a cooperative French-American space program The satellite was designed to gather data from 115 balloons each carrying an instrument package around the earth at an altitude of 38 000 feet On command from Eole the balloons could transmit their information to the satellite which in turn would relay the data to a computer center for analysis The balloons all carried explOsive charges which could detonate on a command from the satellite On the 346th orbit of the satellite a French programmer error caused the Destruct command to be sent to the satellite instead of the Interrogate command The error was discovered quickly but before the instruction could be rescinded the satellite had hurtled over the horizon beyond control b NASA spokesman said I couldn I t tell you wha happened after that sort of chaos broke loose in the station Eole destroyed 72 of the 115 balloons all of those in its path on that orbit NASA officials said the mistake resulted in some procedural changes at Mission Control and the possible demise of one dumb computer programmer James Martin SeC U1'ity Accuraay and Privaay in Computer Systems p 10 Maintenance System E1'1'ors No system can be regarded as secure unless both the hardware and the systems programs are October 78 CRYPTOLOG Page 3 UNCLASSIFIED DOCID 4009815 UNCLASSIFIED designed for security The purpose of system controls is to prevent unacceptable data from being processed and to detect it if jt does enter the system System errors either hardware or software can cause a failure of one or more of the protective features The system should be designed to log all attempts to access data files that failed due to a loss of the proper permission or password management It s up to management to see that the procedures are followed by all I also feel that if everyone would take more care and have more pride in his or her work fewer mistakes would be made I Ihas---pl'epared--a--bibliography---on P L 86-36 P L 86-36 the topia dis-aUB-sed in this- artiale To obtain a aopy mte to CRYPTOLOG P1 01' aal-l-the editor on 5236s U Conal-usion Security is only as strong as its weakest link and the weakest link in the security chain is people The weakness arises partly from the fact that people tend to evade or ignore the standards and partly from the fact that people tend to concentrate on only one thing at a time When there is pressure on the data processing department -- projects going live last-minute modifications emergency maintenance -- observance of standards falls off The value of data which is processed on computers such as social security records and confidential information is immeasurable Consequently protecting equipment and data from unauthorized or inadvertent acts of destruction alteration or misuse is a matter of inestima'ble importance Data processing security is a means of safeguarding hardware software and data against loss from accidental disclosure of data and or modification of data The control of data-base vulnerability is a significant problem in many computer systems There is no method avaiIabl for measuring the quantity or quality of security that may be adequate for a computer In many cases too little attention and too little money is allotted for computer security The effectiveness of a security system depends on the interaction of people within the data processing system The implementation of proper procedures can help to regulate the interaction of personnel and the computersystern thus improving the security -of the system Maintenance of security demands compet nce loyalty and integrity from all personnel connected with the system In addition it requires continuous training for them both in operating procedures and in security measures The purpose of this training is to ensure that each individual recognizes his or her vital role in security practices and does not through familiarity with the system become careless It is my opinion that each individual who is exposed to the computer environment at periodic intervals should familiarize himself or herself with the procedures established by WHO Ql WHOM I P16 Nobody Else ' S orne people's last names have become such household words that we may forget that they have -- or had -- first names This is especially true when two or more names get strung together It's much easier to talk about Lunt Fontanne than to throw in their first names too _ We have listed ten famous pairs on the left below and over on the right there are two columns of first names In each case the names in Column A are those of the first member in each pair and those in Column B are those of the second member If we included Lunt Fontanne Alfred would be in Column A and Lynn in Column B You will note that all three columns are in a phabetical order so just by coincidence a few names are in the right place But we won't tell you how many or which ones 1 Barnum Bailey 2 Currier Ives 3 Dun Bradstreet 4 -Funk WagnaUs S Gilbert Sullivan 6 Lewis Clark 7 Mason - Dixon 8 Sears Roebuck 9 Simon Schuster 10 Taft - Hartley October 78 CRYPTOLOG Page 4 UNCLASSIFIED Column A Column B Charles Isaac K Merriwether Nathaniel Phineas T Richard L Richard W Robert A Robert G William S Adam W Alva C Arthur S Fred A Jr James A James M Jeremiah John M Max L William Answer on pag 18 U DOCID - 4009815 ------ 'fOP SBCRB'f f TMBRA EO 1 4 c P L 86-36 P L October 78 CRYPTOLOG Page 5 tfOP SBCRB'f UMBRA - - - - -_ ----_ _-- i EO 1 4 c P L 86-36 86-36 DOCID - ------------ - E O 1 4 - c 4009815-- P L 86-36 TOP SECRBT UMBRA KNOW YOUR GEOGRAPHY I P L 1 5 ost countries are difficult to recognize when looked at in isolation How many of the countries shown which are regularly in the news can you recognize Countries are not drawn to scale If you can get more than five correct consider yourself at the head of the class 1 7 2 8 3 9 4 10 5 11 6 Answers on page 19 October 78 CRYPTOLOG Page 6 TOP SECRET I TMRRA U 86-36 DOCID 4009815 UNCLASSIFIED he prospect of a postprofessionalization program seems to loom in the future A viable alternative would be to create a continuing system that included all people within a given career field It would mean starting anew and rethinking the scope purpose and objectives of the present system Since most of us now have a stake in the present system we might tend to deem any new system as threatening Thus any changes to the present professionalization program will undoubtedly be gradual and incremental with little likelihood of any radical departure from what exists now A revolution in professionalization cannot occur however unless someone makes the leap to begin the struggle for change The following is my rationale for a continuing system All current professionalization programs at NSA have varying types of requirements with assigned point values given to them Secondly all have a target point at which that final hurdle is completed and professionalization is yours Following professionalization your records are returned to you and you become a statistic There are few pressures on you either to maintain or to improve yourself within your career field from self-motivation and a certain amount of organizational peer pre ure It is the view of this writer that all professionalization programs should have no final hurdles The scope of these programs should be large enough so as to include all people who choose to call themselves a professional There should be requirements within the system to make it necessary for everyone within the field to both maintain and improve their skills The assumption here is that there is always more to learn and time to improve no matter how much you have distinguished yourself within a particu- ' lar career field Secondly professionalization should serve as a measure of a person's skill and worth -- at least from a technical point of view Professionalization should guide people toward achievement and increase their motivation throughout their career All work would then follow the path of the high achievers if the professionalization requirements are skillfully and fairly established and maintained T Professionalization panels should be people oriented 'and serve more as the employee union representative vis-a-vis organizational prerogatives Secondly the panels should get out of all forms of testing and leave this function in the hands of the Training School Within guidelines panels should determine the curriculum necessary for professionalization and then help people achieve their goals With the functional control of their profession the panels could then provide an organization with professional help as needed and remove them when necessary NSA would then become task-oriented The panels should also provide the lateral input to organizations so as to ensure equal career development opportunity Career panels with a larger scope and purpose could ensure functional equality within a given career field and thereby largely elimi nate the promotional aspect of being in the right place at the right time Unfortunately it is now possible for people within the same career field to be hurt more by their timing than by their ability A person with half the ability half the experience and half the whatever-is-required can now be promoted ahead of someone in the same career field who is twice as qualified A more powerful panel would be better able to prevent such inequity A continuing professionalization program which included all people within a particular profession would better help managers to quantify people's strengths and weaknesses It would also enable the Agency to measure statistically the professionalization programs and its people from year to year or whenever necessary Presently no one can quantify the statistical worth of each profession to say whether it is improving or declining in expertise Comp risons should be possible to provide more effective help in recruiting and career self-monitoring Now the process is basically subj ective in respect to the quality of people within given career field It is time for the Agency to become more scientific in its approach to professionalization and increase the level of functional inputs which could help to increase organizational output In order to remove any reader I s fear of what a continuing professionalization process would inVOlve a hypothetical work sheet for evaluating aspirants is presented on the next page October 78 CRYPTOLOG Page 7 UNCLASSIFIED a DOCID 4009815 'UNCLASSIFIED REQUIREMENTS FOR CERTIFICATION BY PROFESS IONALlZATlON PANEL X REFERENCE LEVELS OF EXPERTISE IN RELATED CATEGORIES WITH NUMBER OF POINTS REQUIRED 2 I 750 IGOO 3 2 00 3300 WORK EXPERIENCE TRAINING EDUCATION PERFORMANCE DQCUMENTATION 5 300 6 5200 2000 500 600 500 700 2 00 GOO 750 600 850 TOTAL POINTS ACCUMULATED BY ASPIRANT PANEL AWARDS IN NONRELATED CATEGORIES ASPIRANT CERTIFIABLE AT LEVEL 3 Professionalization awards would be given when a person has accumulated the required number of points for the particular level of expertise For example in the hypothetical instance above 2400 points would be required for professionalization at level 3 The point values given here are of course arbitrary in the real-life situation the point values would be assigned by the Panel and would perhaps be approved by an ad-hoc Panel Comparability Board A person would be able to accumulate the required number of points from various reference levels of expertise in categories related to his field of specialization and also from points awarded for categories pertaining to work experience etc which is not related to his field of specialization The Panel awards for nonrelated categories should be assigned point values and be standard for all aspirants For example a person working in Special Research should be given more points for a college course in international relations than for a course in mathematics The reverse would be true for a Mathematician Education points should be fitted to the professional discipline education and documentation A promotion board looking at this work sheet might wonder why with all his experience the person has not documented his labors for other people's benefit Since a person's reference level of expertise as his career progresses should ideally be as close to the vertical as possible this work sheet might indicate a reordering of some career priorities The problem may rest with the individual or the organization For example the organization may not have allowed the person enough time to document his work Alternatively the person may not write well and may be hesitant to document his work In that case the supervisor could suggest that the person take a technical writing course before zeroing in on his documentation task The advantages of a continuing professionalization system are o It is intended as a method of quantifying a person's career from the beginning to to the end There are no final hurdles which artificially handicap a person o A person could be better identified according to his expertise in different The reference levels of expertise would career fields For example a person help supervisors and aspirants to realize the might be a level-l Collection Operations areas of needed improvement Each person's Officer a level-2 Special Research Anaassets are different and each person's prolyst and a level-3 Traffic Analyst fessional advancement takes a different route o The system would p ovide for complete One person may be long on experience and short quantification of skill levels within a on education and another person may be just career field This would aid management the reverse A glance at the chart would show by increasing the efficiency in recruiting the supervisor and aspirant that if the person staffing transferring and promoting wants to develop his capabilities to the next people and higher level of expertise he should try to o People would be more motivated to improve bring up his totals in categories with low their professional skills totals Otherwise he might prove to be topThe greatest disadvantage of the proposed heavy in one category to the detriment of his is that it makes the present professionscheme overall career development alization system obsolete This would be the largest of all hurdles to overcome To this In the sample work sheet above the aspiwriter however the concept of continuing' rant's accumulated points indicate that he is long on work experience and should probably try professionalization is vastly superior to any program of postprofessionalization to bring up his total points in the area of October 78 CRYPTOLOG Page 8 UNCLASSIFIED DOC I D - -40 ----- - - ---- ----- SECKET P L 86-36 Ix-VV322 When the author of this artiole submitted it to CRYFTOLOG in April he stated that he had deliberately omitted muoh of the more teohnioal detail suoh as how the probabilf t figures are derived but offered to answel' any questions that the published artiole might engender Unfo 'tunately that offer no longel' applies sinoe the author l'esigned in late June 1978 Questions may hoUJever be direoted to his former assooiates in W322 on 3764s Colleotion Editor tarticle in the April 1978 CRYPTOLOG We Gotta Accentuate the Negative pointed out the problems that exist with intelligence efforts that yield negative results The challenge is to make the most of such results and produce negative but useful intelligence This article will attem t to describe a method develo ed b G October 78 CRYPTOLOG Page 9 EO 1 4 c P L 86-36 EO 1 4 c P L 86-36 8R RRT WAil9bE Jill ee tHff EURllI'diNEbS DOClD EO 1 4 c P L 86-36 4009815 NILLIT REALLV DO THE 08 accelerating spiral of technology catches up At this time one must either put the equipment to new visionary use or plan for a new purchase 1'2141 ' ' na day of spiraling inflation and spiraling computer-related technology the staid method of selecting the best equipment may be impractical If we select an equipment only for its ability to perform we will more than likely ignore several other crucial criteria If criteria such as those I have in mind are ignored our future operations may be seriously hampered I It was once said that he' who has not learned from his past mistakes is c ndemned to reli e them And relive them he w1ll unless flex1b1lity of technology is one prime consideratio in selecting an equipment What degree of flex1bility is built into the technology of a contemplated purchase The knowledge of built-in flexibility would certainly be a comforting thought when -- and it is is inevitable -- the On the other hand what we buy we must support Several questions should surface in this area These questions center around the con trac to r ' al iJityandwiUingnes s tofurnish aconsta t stream of technical support for an equipment he wants to sell So we ask does the contractor have a maintenance team available Can we easily identify and obtain replacement parts Are good technical manuals available including logic drawings and technical changes Does the contractor give training courses on his equipment Does he have both the ability and the willingness to develop software and drawings specifically for our needs It is this type of thinking that conveys to the contractor the message that the equipment he sells in June he must support in December It is needless to say that all these considerations are necessary during the preliminary planning stages Otherwise we may abruptly discover that the ideal equipment may fall far short for extended use and be doomed to a very short life cycle lU October 78 CRYPTOLOG Pag 10 SRCRE'f ItI ql'cE fA selmlT EIWlIlHS Q lb P L 86-36 DOCID 4009815 CONPIBHN'fIAJ EO 1 4 c P L 86-36 YOU CAN'T TELL THE WHEATFROM rHE CHAFF WITHOUT A PROGRAM J Gurln as ----- ---- o -October 78 o CRYPTOLOG Pa 11 C9HFilliNTIAI IhOdH3bE 'IIA EURSMHi't' EURI1 AWIEhS SllhY DOCIO 4009815 UNCLASSIFIED REDUCTION RATIOS IN MICROGRAPHICS Don Snow V1 DDO MICIOOU HICS COOIDINATOI III1III1III 4 6 ake a sheet of letter-size paper Fold the bottom edge up to the top Now fold the right edge over to the left edge You have reduced the dimensions of the original sheet by two or 2X but you have reduced the area by four In micrographics when we talk of reduction ratios we're describing the number of times the dimensions of the original material are reduced So a letter-size page at a reduction ratio of 24X sometimes expressed as 24 1 would yield an image one-third of an inch wide by just under a half an inch high As many as 98 such images can fit on one 4-by-6 inch microfiche stead of being 14 by 11 inches is about threequarters that size That still affords a very legible projection but the reader does not require as much desk space There are a few portable or attache case -type readers but they lack the quality of their larger brothers The two standard reduction ratios 24X and 48X involve one photographic step In the case of source document conversion at 24X pages are fed manually into a machine which then photographs either one or both sides automatically and repositions the microfiche for the next page image Source document conversion to microfiche is best done at 24X reduction' h wever computer-output-microfilm COM 'may use eIther 24X or 48X reduction depending on the desires of the user The dimensions of a standard computer printout page are 14 x 11 inches Again using a reduction ratio of 24X it is possible to put There are greater reduction ratios in today's 63 such images on one microfiche When a desired page image is positioned in a reader it is en- s ate of the art In a previous article The BIble and the Washington Monument CRYPTOLOG larged to the original dimensions of the page -14 by 11 inches thus the alphanumeric charac- S ptemb r 1976 I described a 2-inch square of ters on the screen are exactly the same size as fIlm WhICh contains all 1245 pages of the Bible That was done using two photographic steps they would be on a printout page each page was shot at a reduction ratio of l5X' Thus far we've mentioned one of the two DoD then four page images were grouped and shot at' standard reduction ratios The other is 48X or a ratio of lOX The final product or ultra48 1 This allows a greater compaction of m - ' fiche contained each page image at an effecterial to the point where l4-by-ll printout tive reduction ratio of 150X pages can be shrunk so that 270 images will fit Perhaps the greatest reduction ratio in use o one microfiche Or if you're using lettertoday is 2l0X another two-step process which SIze originals 420 images will fit on one microfiche And by using the proper lens you roduces an ultra-strip measuring 1 by 7 can blow back one of those tiny images to its Inches Such a strip can contain up to 2 000 original dimensions The microfiche readers in pages And by using a keyboard and info ma- use here at the Agency are equipped with lenses tion displayed on a cathode-ray tube by a minicomputer a user can retrieve and project any to handle both 24X and 48X reduction ratios one of those pages in 4 seconds But that's another story In recent years there has been a trend toward making readers less bulky so that a screen inOctober 78 CRYPTOLOG Page 12 UNCLASSIFIED DOCID 4009815 UNCLASSIFIED BACK TO SQUARE ONEI T he trend toward microforms in all categories of publications seems to be facing a setback The publishers of the Ameri- can Journal of Computational Linguistics which has been sponsored since 1974 by the Association for Computational Linguistics in a 20X microfiche format have recently polled the membership about format preferences Although previous membership samples have indicated support for the compact format this latest request for direction from the membership is couched as an appeal for a change to a more traditional format such as used by the Communications of the ACM and the various IEEE Transactions The reason To increase the readership and consequently the rate of submission of technical articles Production costs would approximate those currently sustained and it would even be possible to produce the microfiche too for very little more P L 86-36 P L 8 6- 3 6 Whatever the eventual outcome for AJCL the apparent low readership seems to be attributable to the alien format Much as I have supported the use of microforms for active files and other uses I must confess my own response to AJCL microfiches has been lukewarm Although I like being able to keep a complete set of copies in a 4x6 card file I just can't curl up in an easy chair with a microfiche reader for an evening of professional enlightenment Is there a solution that the human factors engineers have missed U OVERHEARD WHILE STANDING IN THE BURN-BAS lINE for initiating a new CRYPTOLOG department I don't know what is The quotes are If they realty want that many linguists around here why doesn't the Director hold his staff meetings in Russian Working all day long on that scope isn't as bad as I thought I keep a book nearby and whenever the system goes down there isn't anything to do so I read It's all very uplifting I'm going to read War and Peace The guy beside De is studying law Have you overheard anything n two days in a rowl Isub mitted to the editor a quote that had been overheard while standing in the burn-bag line If that isn't enough reason O interesting while standing in the burn-bag line If so send it along to this department Sign your name or keep it anonymous it's up to you Hint to would-be authors Keeping an ear open while you're standing in the line yourself or keepinganegeopenwhilel'eading this department might give you an idea for a CRYPTOLOG artic le -such as Is Doum-Time Excessive U October 78 CRYPTOLOG Page 13 UNCLASSIFIED DOClO 4009815 UNCLASSIFIED 'IEDI'NG THE GERMANS MISINFORMATION - P W fiLlY Review of of the handlers of of the highest-level intelligence of the war He tells without bitterness how his meteoric rise was not matched with corresponding promotion and he remained a ir John Masterman in his The Double Lieutenant Commander for he whole of the war Cross System in the War of 1939 to 1945 Being missed out for promotion was the fate of published in 1972 told how German spies many cryptanalysts also With his wife in the were captured and then used to feed United States he was still forbidden to stay one false information to the Germans At that time day over the week demanded by his discussions the ex istence of the cipher systems classified wi th his American opposite nUDibersin Washington Ultra which were broken by the British at Readers might also like to picture security Bletchley were still under wraps and Sir John in those days Montagu had to travel between had to use only information which came from departments with top secret papers and he reother sources Now Ewen Montagu a distinguished p orts that thi s put him in a dilemma He judge has produced the complement to Sir John's writes This problem I solved by fixing a large book in his recollection of double-cross and the metal basket-type carrier to the front of the manner in which through Ultra success or fail- cycle To this I chained the brief-case when it ure could be judged Some who worked at Bletch- was parked There was some doubt about the seley deplore the publication of Winterbotham's curity of the operation for such 'hot' documents The Ultra Secret but readers will be enchanted but I managed to get permission to adopt it as by Ewen Montagu's book because he is able to use long as I always wore a shoulder-holster and all the secret intelligence gained 'through 111tra automatic pistol Many readers will remember the startling Montagu became one of the very few to be let story written by Montagu called The Man Who in on Ultra he attended the famous XX CommitNever Was about floating a body bearing 'war tee which met weekly to decide on deceptions to plans into Spain Operation Mincemeat At the 'b tried and therefore needed the information time Montagu could not tell his public that the He had two roles supplier of information to the fact that the Germans bit the whole deception S ices and producer of hicken feed ormisw was ascertained from Ultra 'messages Now he i o rmation to the enemy He t 11s story after t reats his readers to a tour de-force on the way story' all with undoubted authenticity Because the Germans were fed with much misinformation of peration Mincemeat Hitler kept forces' in the through German spies'who were turned round wrong place for over a year while Italy was He names agents and tells how when caugnt they being nvaded still believing that the main accepted the role of double agent and played thrus twould be on the west coast of Greece and supremely well and fairly on the Allies' behalf Sardin1a When the invasion of Normandy was imMoreover through Ultra the British could check lIIinent thq Gel'lllails had tociefend t he Pas de on the manner in which the Germans accepted theC lais'long after t heAlli $ hada firm hold in information and whether an agent had been blown France' sillply becatlse MO t P s team had reported Montagu recounts several astonishing and ex citing ' great build-up'ln the Dover area He had stories of narrow escapes and surprisingly retrouble when there was a moral issue Should he cords that in spite of the steps needed to get eport to the Germans that the U-boat commanders false information to the Germans the stories were the German Navy's heroes were traitors No accepted without much question and one or two But when flying bombs fell in London should he star agents were awarded the Iron Cross I Perhaps get the Germans to believe they were overshootthe most astonishing statement is that throughout ing Yes because fewer lives would be lost to the war every agent'was caught within a few days bombs falling short of densely populated London of arrival probably because of England's being One might wonder how the Germans were tricked a tight little island so often Well not all 'the information was Ewen Montagu had been called to the-bar in wrong Quite often the Alfies 'refeased perfectly 1924 When war'broke out he joined the Navy valid information -- just a little too late for and was assigned to low-level intelligence but any great value to be gained but the Germans such was his brilliaJ lce that he soon became one could prove the information correct and the agent therefore reliable In this book Ewen' ritagu has produced the' See CRYPTOLOG December 1975 Winterbotham's 'The Ultra Secret' -- Three Views A Personal most exciting of all the Ultra revelations to Comment by Brigadier John H Tiltman Ultra appear so far Was Secret Weapon That Helped Defeat Nazis by Du roing WO'l'ld War II the author was an offiaer P W Filby and Mum's Still the Word by in Miti IntetUgenoeat Bletehley Paul R Hutchinson S October 78 o CRYPTOLOG o Page 14 UNCLASSIFIED DOCID 4009815 UNCLASSIFIED W p o L o _ _ _ _ _ _ _---llacldi O 8 6- 3 6 AND A YOU BETTA K CRYPTOLOG you finally got my attention At the risk6f falling for a belated c ryptologic April Fool's g g I really have to comment onl I 1 '1--- ' 1larticle Uncle-a Sam WaRtsa You CRYPTOLOG May 1978 At the end of his article I loffers the opinion that when recruiting linguists high ambition in the applicant should be viewed as a criterion for nonselection He goes onto state that the problem with highly motivated people is that they expect and emotionally require high productivity to be rewarded and government service is not set uptogive rewards for productivity I IthEm offers us a formula for recruiting linguis ts If the object is to recruit career linguists no more than a moderate amount of ambition or motivation should be allowed in 8 recruit Assumingl lis really serious and I applaud his motivation in undertaking this ambitious article I strongly differ with his viewpoint From time to time in the past we've been through this hire he poor and keep 'em without shoes syndrome at NSA The problem is that people's aspirations change as they mature and after they have satisfied their basic needs get a job and eat regularly Inevitably some percent of the supposedly unmotivated work force begins to show dangerous inclinations toward greater responsibility take new and nonstandard approaches to their jobs and horrors begin sinfully lusting after positions where they can influence how and why things are done rather than to gratefully accept their designated place in the system When you think about it Thomas Edison turned out to have been a pretty poor selection when he was hired as a newsboy True he eventually revolutionized communications rather than just delivering them but you can bet that the guys down at the newspaper's recruiting department didn't see it that way and redoubled their efforts to find a replacement who would be satisfied with just delivering the damn newspapers Let's face it our future key people will not have been hired in response to a want-ad for Key People Wanted but will come from among HAVE IYIOTI-VAYSHI l Is perhaps even to include a maverick linguist or two to the eternal anguish of those who recruitedthem as linguists With a little luck they'll have the moxie to solve the linguist problem that has eluded the rest of us for so long I also question the validity ofl contention that government service l- s n 'o -t- s - e lt -- up to give rewards for productivity In my view that depends on one's perception of rewards and in addition to the bucks involved in rewarding people important there are other intangibles which weigh heavily in job satisfaction and the resulting work force stability Since I'm apparently the only person in the cryptologic community not previously embroiled in the linguist issue here's my two cents' worth o Hire the most motivated most ambitious linguists you can find o Tell them up front about the problems of day-to-day language work and the riptide of opinion that prevails on how to acquire and keep a linguistic work force o Challenge them to solve our problems o Listen to what they say after they have some experience o Enrich their jobs encouraging them to stay in the field Pay them fairly o Disregard all they say upon entry about being satisfied doing the exact job they're hired to do My theory is that at recruitment interviews people tend to tell us what they think we want to hear Interviewer And do you think you'll like the work Applicant Actually ever since I was a small child I've had an insatiable desire to sit at a desk and translate all day When all my friends were out playing I used to translate the labels off of olive-oil cans' Virtually all will profess a love of doing the job we have in mind for them vowing no October 78 CRYPTOLOG Page 15 UNCLASSIFIED DOCID 4009815 8BEURRB'f' aspirations for other things Once hired however a bunch of them will begin their private guerrilla campaign for the Directorate and for them their entry job is only a first step Great To hire unambitious workers as a matter of policy as suggests mismtougive validity to the claim that we have in fact become a true old-line bureaucracy and are proud of it Our goal has got to be to make our linguists' work more interesting rewarding and challenging and to make an NSA linguistic career a desirable profession rather than a feudal system where the linguist vassals gratefully accept their piecework assignments From the symptoms evident to me a nonlinguist it seems that we have not managed our language people among others and their tasks in a way that works both for linguists and management The danger is that we will default to a mode of the self-fulfilling prophecy where we hire unambitious people who will then indeed rise to the challenge and become and remain truly mediocre 'P L 86-36 P S Since writing the above I have had second tho ughts about a work force composed of unmotivat fd unambitious people I have been able to 'peate SeV ral linguists who seem to meetl' standards for low motivation and lack of ambition and would make ideal experimental subjects to test his thesis Un fortunately I have been unable to get them to take the initiative and get their applications in fortconsideration u News of the Communications Analysis Association By W E S fit EO 1 4 c f you have ever recovered a frequency work at Arlington Hall Station Hired as a P L 86-36 rota or broken an authenticator system Communications Clerk he performed traffic or figured out how an address table works ana ysis onl then you have been doing Crypto-TA printer and'-- Mo r-s-e -a-n-d s-e-r-v-e 'd a-s-a-s-e-c-t io-n- l whether you realized it or not reporter through November 1956 He joined the Most analysts need to understand enough Office of Training as a TA Instructor for a about this subject so they can 1 solve t o-year tour and tayed 13 years After their own simple problems as they come across teaching Basic Radio Communications and Traffic them and 2 recognize when they need Analysis for some 4 years he then developed specialist help Andthere area few people and taught Traffic Analysis T echnical Reportaround who have the interest' and abi1 ty togeting and SIGINT Repordng9ourses He served deeper into the specialist aspects of the as the Training advisor t the'TA ta r e r subject If either of these descriptions seems Panel when it was first forbi$Ci and was selected to fi t you then you ought to get iil touch with ' to serve as the NSA Pac ficStaff Training the CTASIG Crypto-TA Special Interest Group Representative Prior to his departure for and find out what they can do for you Among the overseas assignment in the summer of 1967 the people you might contact are I he served as the Executive for an ad-hoc gr0tlP P L 86-36 I J charged with determining whetheroI'notE duca and Fred Mason tion and Training sholJldbe onsidered a sareer _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _jGT field withcriteriaf0I'Professionali ation estabIished l arlwas instrumental in estabA follow-up discussion on the subject lishiTlgthat set of criteria before leaving Interstellar SIGINT wasc Ilducted on 14 AugfaT the NSAPAC job After 2 years in Japan ust 1978 forsollle20CAA members with many trips thr ughoutthe Far E st he I hief W16 gave a short recap of the presentation he had given in the NSA Auditorium returned to theNCS National Cryptologic School Sta ffat For tMeade and was assigned to on 28 June 1978 and then entertained questions investigat e the possibility of setting up a The session lasted about one hour -- members tI'aiillng course for troopers on their way to -present came prepared with very interesting Korea similar to the TA operational training questions A transcript of the 28 June talk given to troops on their way to Vietnam is now being edited for publication in a future He then assumed the responsiblity as Chief of issue of CRYPTOLOG the TA Training Department of NeS for approxi U mately one year before accepting the appointment as Executive to the TA Career Panel in May 1970 Meet our Program ChaiP an His term expired in May 1973 but was extended ha been in the cr tologic for one year In May 1974 he transferred to R221 business since 1947 After 5 years WIth the HF Systems Architecture Project Management Naval Security element in Washington D C as In July 1976 he was appointed TA Executive agaIn a Communications Technician he converted to and currently serves in that capacity 5 Q civilian status in February 1952 and went to I I I I I October 78 CRYPTOLOG Page 16 SRenT L DOCID 4009815 86-36 JONFIBBN'fIA MORE B S l Before SPELLMAN L -- IG03 ' Another remembro nce of projects past prompted by Jack Gurin' s Never Again CRYFTOLOG June 1978 and A SaZemme's foUow-up to it I Remember SPELLMAN CRYPTOLOG JuZ1 J'-August 1978 1 nce upon a time 1956 a tender soul me was employed by the Agency which was looking for someone with knowledge of languages area studies and intelligence background other than COMINT And so I became a minion in the TNG Training fiefdom of the late lamented Shelby P where books on desks had to be aligned by size and not by content from left to right where suit coats had to be worn anytime you left your desk for a trip to the you-know-what and where everything inclUding thoughts was hierarchically structured But times were changing and even Camelot looked outwards And one day someone decided that the Agency should look into management theory So soon TNG hired a couple of fellows who would dream up management courses Their room was next to mine in fact their entrance was through my office so I overheard maniof the s J1'lllising$ that 6ventually found their way into the first trial and protoseminars -- we next door laughed at the absurdity of most of their notions but of course such endeavors have long since been anointed as the Holy Grail and are believed in by people who look askance at ESP flying saucers and all of Serendip About that time the concept of brainstorming was trying out its wings in the world and of course the management people picked up on that gimmick Thus is came to pass in the TNG fiefdom that the Grand Vizier about that same time Lambros Callimahos took upon himself the title of Guru promulgated an edict to the effect that on Friday mornings assembled his minions would brainstorm in his presence His minion of the Irish persuasion soon asked for support O from his own 511hm jpjoD5 I myself 50 mjahtily djd and others come up with seed ideas Jnd other things for brainstorming input One of these ideas for the brainstorming sessions had to do with -- would you believe -- stenotypy and possible ways of having tran- scription done by people with less than the five years' experience with spoken Russian that is required to do professional-level transcription Whether that idea was ever thrown into the Friday brainstorming hopper or not I do not know But herewith is the exegesis of that particular ploy In 1949-1950 I had been an English-language teacher at the Lyc e Henri IV in Paris One of my students there was also studying stenotypy at the Grandjean Grand Jean Institute or was it the Grandjean system I don't recall Anyway that student and a couple of other stenotypy students asked me to help them -- they needed American speech they could practice their stenotypy on I obliged One of the students was French-born but the child of Polish refugees She knew French Polish some German and was perfecting he English Students of the Grandjean system in Paris were of very diverse backgrounds coming from allover Europe and from the various French colonies Therefore by design the Grandjean system as taught there was deeply influenced by phonological theory My little Polish student found it easy to use her stenotype machine to record Polish and English as well as French But back to Camelot During my early years with NSA I lived at Hartnett Hall where we had a TV room At 6 p m every night two of the Hartnett Hall residents one of whom was taking stenotypy at some local school would take their seats stenotype machines in hand and practice transcribing the news broadcasts One night there was possibly up to a minute's worth of foreign language on the news probably De Gaulle and I noted that one of the stenotypists continued to stenotype without a break After the news broadcast was over I remarked to him that I didn't know that he knew French and his reply was that he didn't -- he was justpracticiIlgthe sounds he heard Heran through his roll of tape until he found the proper spot and then pronounced what he had on paper I could understand it but he ouldn't It was sort of like the old gimmick of asking a French speaker to read aloud the nonsense phrase pas de leur Rhone que nous in front P L October 78 CRYPTOLOG Page 17 JONPIBt N IAL p -I ' i 86-36 DOCID 4009815 CONFI9BN'flll of nonspeakers of French and have the latter burst forth in laughter while the French speaker is totally nonplussed since the phrase makes no sense -- after all he is filtering the words and sounds through the French phonemi pattern and doesn't realize that his audience is filtering the sounds through the English phonemic pattern and is hearing Paddle your own c noe Q E D I had learned that i t is posSIble for a person to transcribe a little bit of a foreign language he doesn't even know Note This was a highly trained stenotypist who was already working for the D C court system As a person with a degree in Chinese from Harvard and a Master's in French from Middlebury plus experience as an announcer in Spanis and Portuguese on WLW short wave in Cincinnati nd an immediate background as an' interrogator' In German for the USAREUR Intelligence Center I should have known that my first major assignmen at the Agency would be the teaching of POlISh That meant that we hired the Georgetown University Polish instructor as a contract employee of the Sanz School of Languages to get my feet wet and bring me up to speed to take courses from him at Georgetown The environment at the Georgetown School of Languages and Linguistics plus the heady atmosphere of the PATA Promotion and Training Agreement Program in TNG in which the students were introduced to the concepts of linguistics using the Gleason textbook soon got me into formal linguistics And thus the stage was set Then came brainstol'lBing The concept as I came up with it was a$ fol lows Some high-school students would take typing and stenography It should be poss le' to selects me peop efrolRthat populatioI cwho have an aptItude for language and teach them stenotypy and drill them in the sounds of phonology as taught in beginning college courses in Phonetics and Phonemics using the texts of the Summer Institute of Linguistics or other such materials Afterwards they would be taught the ds of Russian The second step would be to take a number of open-minded Russian linguists and teach them stenotypy The purpose was not to turn the Russian linguists into stenotypists but to enable them to read the stenotype tapes made by the non-Russian-speaking clerical specialists Back in 1956 we had a Russian-language manpower shortage we still have one except that now' it probably ought to be called a personpower shortage -- remember that it takes a degree in Russian and then up to 5 years' experience to become a truly professional transcriber -- but the stenotypy concept could possibly train people in 6 months or so to put onto stenotype tape information which a skilled linguist could scan for relevance and importance Then the decision could be made to turn the tape over to a transcriber In other words my concept was that this was a method of making it possible to P L 86-36 scan a far larger volume of material by eye than can be done on a I 1 basis by ear The key and salient points of this two-step concept were o the careful selection of highly motivated clerical people and the training of them first in stenotypy and general phonology and then specifically in the phonology of Russ an and o the providing of hand-in-han training in stenotypy to some senior 8ian linguists for scanning purposes Never did I say that G I 's selected in the military manner to learn Russian at a 6-month course in Monterey should simultan ously be taught stenotypy and become stenotypy transcribers nor that a Cyrillic keyboard would be needed nor that this could be an i put into machine translation So you can iJllaginemy surprise one day in 1960 when I learned what was being asked o I stated to her then and there that It sounded like my original concept of a few years earlier but one which had been brainstormed and confounded until there was no possi le way it could succeed If they do want to try again with the stenotypy concept some day I'd like to implore them to please-oh-please do it right I This means the proper selection of two types of individuals the proper in-depth linguistic training of the stenotypists and matched paired stenotypy training of the clerical people and oftheseanner ltnguists And also have ataastone of the lIOJ itofsanddevelOpers of the prog raIR beasci_tiiic liJlgltiftWith the requisite lmow1edgeof thi i'Osslblepitfalls and difficulties I ' stiflconvinced that a proper tria1 could well have positive results I -- -- 2 ANSWEI TO VVHO VVHOM PAGE 4 1- 2 3 4 5 6 7 8 9 10 Phineas T Nathaniel Robert C Isaac K William S Merriwether Charles Richard W Richard L Robert A BARNUM James A CURRIER James M DUN John M FUNK Adam W GILBERT Arthur S LEWIS William MASON Jeremiah SEARS Alva C SIMON Max L TAFT Fred A BAILEY IVES BRADSTREET WAGNALLS SULLIVAN CLARK DIXON ROEBUCK SCHUSTER HARTLEY Jr The only two names in the right row And Harry says If you don't believe me you can look it up in your Isaac K Adam W I October 78 CRYPTOLOG Page 18 -CONFIBBN tAL U DOCID 4009815 FeR eFFIEURIAL ijSE 8NL i ' 6 P l wondered then and still would like to know about the freauency in I I J-r e'l f e r e n -c -e -s -- t o ma ' il- e t r' 'a'j f'j f ic a n a lr y s t s - -iip e r haps some scholar has made a study P L Vera Filby E41 86-36 To the Editor CRYPTOLOG larticle in the July-August CRYPTOLOG Ye Gads Another Country Trigraph System ends by asking if there isn't someone who can solve the problem he brings up -- proliferation of country trigraphs I suspect that he is already aware that such a group already exists -- the NSA Data Standards Center P13D What he may not be aware of is that our group has no teeth so we cannot force anyone to follow our lead The data standards problem that Jack brings out is actually only one of the many that plague this agency as well as the whole U S government As things stand now we are not really in a position to know of many such problems until they are full-blown Even the Senior Data Representatives who are supposed to see all jobs before they are processed too often find that they have been bypassed Data files and data bases are designed and established long before we are brought into the picture and any attempts we make to correct errors are either ignored or circumvented We have no power to make them change and they know that U To the Editor CRYPTOLOG After several long hours of struggling to complete the July-August NSA-Crostic distributed across many lunch periods I was somewhat disappointed when I finally read the text that I had recovered Innocent as it may have seemed to Mr Williams I found it to be in extremely poor taste to perpetuate through this puzzle a bit of humor regarding female Traffic Analysts Although I now work as a Computer Systems Analyst I have also worked as a Traffic Analyst and still consider myself to be a member of that profession The quotation in the crostic is the type of statement that promotes basic prejudices that still exist in the Agency because they exist in our society I will laugh at jokes about female Traffic Analysts only when the opportunities for professional women are the same as they are for men and when the rate of women promoted is equal to that of men promoted and when the number of women in management positions is proportionate to the number of men in management positions Furthermore until that time the CRYPTOLOG will be doing an injustice to its professional quality by publishing such backhanded affronts to female employees As the NDSC was first conceived it was supposed to have the capability for monitoring and investigating jobs to see if they varied from the norm The NDSC would also have had the T323 authority to take appropriate action if the initiator failed to conform to standards For example we would have been able to order the processing to be discontinued or we could have ANSWERS TO KNOW y'OUR GEOGRAPHY p 8 denied access to computer systems to those pro1 Libya 7 Norway grammers who were habitual offenders 2 Cuba 8 Jordan 3 Sorrr zUa Yes Jack there i8 a standards center but 9 Thailand 4 Veneauela until we are authorized to deny computer use to 10 Panama 5 Austria those who refuse to accept Agency standards we 11 Spain 6 United Kingdom U can have little effect Mark T Pattie Jr Solution to NSA-Cro8ti No 17 Chief NSA Data StanBy A J 5 CRYPTOLOG September 1978 dards Center tf f V'MofAUV -lJ' I IU The NSA Intern -------------------Program NSA Ct'yptologic Spect'l'Wl1 To the Editor CRYPTOLOG Fall 1977 Vol No 4 Dave Williams' NSA-Crostic No o 16 in the The intern program s beg ln to provide an July-August 1978 CRYPTOLOG remin4ed ae of the accelerated system through which new college Quest i on I meant to ask when I r e a d some lnsa nce t s experIence d _n______ lP 'Sd U8 tes and In Letterto the Editor in the July 1977 1 d t t f onal emp oyeel were tr 1ne 0 lI ee pro eSSl CRYPTOLOG from whichthepuzz'1eqUQ t a t i o I l _ level mannihgr quirements ' ' taken The quotation included the observa lon U IL - I i I rof 1 I I __ _ -- r I J oooooooo Letters W tlu Ediwr I 6 October 78 CRYPTOLOG Page 19 FQR 9FFleIAl 9E 6NLY I P L 86-36 DOCID 4009815 UNCLASSIFIED N8A-cro tlc No 18 By guest NSA-crostician David H Williams P16 P L 86-36 The quotation on the next page was taken from he published work of an NSAer The first 'Letters of the WORDS speU out the author's name and the tit'Le of the work DEFINITIONS October 78 CRYPTOLOG Page 20 UNCLASSIFIED DOCID 4009815 PI-Oct 78-53-26638 UNCLASSIFIED October 78 CRYPTOLOG Page 21 UNCLASSIFIED P L 86-36 DOCID 4009815 1il Ii III ---- 1----1 1 Q l L--_ _ I THIS DQCYM8NT CQNT INS CQD8WORB MA'I'RRIAIJ This document is from the holdings of The National Security Archive Suite 701 Gelman Library The George Washington University 2130 H Street NW Washington D C 20037 Phone 202 994-7000 Fax 202 994-7005 nsarchiv@gwu edu