980 m 'ilVmm k 5 rnWV'il GBPDlD 5m rUJ V l l15UJ l l15 l l OO15 fil15 OO l1U1llJfil P L 86-36 - L CODE WORD OR COMINT CHANNELS 86-36 I o 1 TA HANDMAIDEN OF CA o oo o o o o Frederic O Mason oo oooooo o 3 HORRAY FOR PMD' S o o o oooo ooo 1 ARE WE WASTING LINGUISTIC TIME 6 Mary Roberta Irwin o 7 LETTERS TO THE EDITOR oo o o o o o oo oo o oooooo o 11 'fillS BeetiMBN'f' eeN'fAINS eeBB teRB MA'fBRIAh 'l'nss lellll 9IRPiS PiS ' taa a HIe_pt ' 0 6B8 Be 1165 tJateiOi the_silT lI s Pistil tis II tile QI'i'- te Declassified and Approved for Release by NSA on '10-'1 '1- 20'1 2 pursuant to E O '135 26 vl DR Case # 54778 DOCID 4009720 Published Monthly by Pl Techni uesand Standards for the Personnel of Operations VOL II NO 5 MAY 1975 WILLIAM LUTWINIAK PUBLISHER BOARD OF EDITORS Editor in Chief o Doris Miller 5642s Collection o 1 P L 1 3571s Cryptanalysis o 1 fB02Ss Language o o Emery W Tetrault 52365 Machine Support o 1 Special Research 1 33215 Vera R Filby 7119s For individual subscriptions send name and organizational designator to CRYPTOLOG PI 86-36 DOClD 4009720 TOP SBEURRBT UMBRA P L CODEWORD OR COMINT CHANNELS To the average person the words AKPn AnOPPHTON An WPHTON and EMnIITETTIKON are just three Greek terms Translated these terms are TOP SECRET SECRET and CONFIDENTIAL but unfortunately to many NSAers they are still Greek We in NSA know how to handle papers and documents with the various classifications affixed our pro lem is in determining the classification of papers that we ourselves originate One of the major classification difficulties is determining whether certain information belongs in a COMINT category that is requires a Codeword on each page or whether it must merely be kept in COMINT channels The definition of COMINT Codeword is information that if compromised could cause damage to national security and specifically to COMINT activi ties Any paper for instance that quotes a COMINT source directly or indirectly requires a Codeword b cause it reveals that we are successfully reading certain traffic--information that if compromised would certainly jeopardize th t traffic Such information requires handling that affords thehighest degree of security protection Of course there are various degrees of damage and of protection depending on whether it is COMINT Category I II or III COMINT Channels information is that which though less revealing must still be limited to COMINT-cleared personnel because it contains information related to COMINT agencies or activities To aid in determining which group our paper falls in we can look at the rules and guidance in the NSA CSS Classification Manual and USSID 3 These guidelines often provide just the information we need However when we cannot find a specific rule or guide to solve our classification problem we must use the general definitions' and good old-fashioned judgment Some hints in the judgment area are Very sensitive SIGINT plans and operations sometimes require greater protection and carry a COMINT Codeword classification just for that reason If the document reveals a degree of success or progress in the production of COMINT or VI a sophisticated COMINT technique a greater protection than COMINT Channels Only is required Information classified by another agency or government may carry a higher classification than that we would have used 'but we are required to apply the same classification as the originator As a Classification Advisory Officer I am often called upon to assist in determining whether certain information is Codeword or COMINT Channels Itis not always a clear and easy decision I use all of the available guidelines and judgment factors mentioned above When this does not give me the proper classification I approach the problem in another way I work on the supposition that the material I am trying to classify is compromised and falls into the hands of the enemy With this information what countermeasures can he the enemy take to insure that such information will not be available to NSA again If the exact time date place and means of acquisition are available from the information he can take extensive measures to insure that such information will not be communicated in the same manner again He can change callsign systems frequency systems schedules and cryptosystems and impose strict limitations on chatter usage and he can do so at a particular transmission site When your information fulfills some or all of the criteria above it can easily be determined to be Codeword On the other hand if the paper gives general information related to the COMINT effort but does not include specifics of a transmission or actual transmission procedures give the enemy information leading to specific countermeasures it requires only COMINT Channels protection or This can be amplified by way of example The following problems are fictitious and are used for illustration only Mr Charles Great is recommended for the Meritorious Civilian Service Award for his outstanding achievements while assigned to the National Security Agency Central Security Service from December 6 1970 to December 9 1974 During this time he worked May 75 CRYPTOLOG Page I TOP SECRET t JMHRA 86-36 mOl 4 c DOCID 4 aa 97 01lli 36 1'QP 8 CK T Ul fHKl ooo oo L _ ---ll He guided the field sites in their collection efforts and was very helpful in solving problems that existed in their processing and reporting efforts This is obviously COMINT Chllnnels It tells the enemy that we intercented nrocessed and renortedl This is obviously Codeword It identifies a unit which sent a message on 15 December and relates the text of the messaoe T 1 I V 86-36 tFOP SHURHtF HMHKA EO 1 4 d 'fSP SEERE'f lRlBfbtt P L May 75 CRYPTOLOG Page 0 1 4 c DOCID 009'720 TOP SECRE'I' UMBR2 _ TA HANDMAIDEN OF CA FREDERIC O MASON JR Pil as an anaI Traffic analysis was created lytic function in order to organize cOllectionl and to maintain continuity for support of crypt- analysis In a broad sense it still does that but it has also achieved a value of its own __ Before TA was fragmented into many career fields for reasons which escape me it was everything in COMINT which wasn't CA or--more properly--it was a subset of CA employing many CA techniques but against a different body of data The essential differences between CA and TA are twofold Generally the two skills are inversely important if CA is very successful TA is apt to be little valued though very complete if CA is stymied then any TA achievement is valued highly In practice the line between the two is sharp since TA normally does many trivial tanalytic jobs are then both similar and in- May 75 CRYPTOLOG Page 3 TOP S CR T UMBR'a EO 1 4 c P L 86-36 DOCID 4009720 --80 1 4 c L 86-36 Ttl' SECRET UMBRA t May 75 CRYPTOLOG Page 4 'fOP SIUJRB'f HMBRA DOCID 1 4 c L 86-36 4009720 TOP 8BSRBT HMBRA If this sort of exercise entertains you we can do some more Let us know May 75 CRYPTOLOG Page 5 TOP 8BSRBT UMBRA DOCID 4009720 ooo HOORAY FOR PMD'S 09 My first impression of Program Management Directives was that the whole process placed an ------------------------------------ inappropriate emphasis on planning As a result I think the implications of a time-phased there appeared to be undue staff involvement in action plan in the points brought out above things that had previously been the exclusive speak for themselves For the first time we have domain of line organizations a uniform rational system for exploring on a More recent experience with a specific PMD timely basis the implications of any given move dealing with the phaseout of an overseas site changed my views radically We in M have for years suffered in varying degrees of silence about not being notified reThe draft PMD was given to me with a note saying I can see considerable M involvement garding operational activities until they were My immediate reaction was To close out a site at a point where only costly super-priority action on our part could provide the necessary You've got to be kidding After a discussion support to successfully complete the operation with the Program Manager however several Now under the PMD system we are participating things began to come into better focus at the earl'iest stage in each proj ect IncreasFirst the Program Manager asked for an ingly we are involved in detailed planning of inventory of equipment to include current value events beyond the immediate future This is size in cubic feet and weight He indicated and should be a challenge We will undoubtedly that he needed this information to develop an find that it is easier to talk about planning emergency evacuation plan He had already disthan to do it I think however we will find patched crates for key items which were quickly that it is more stimulating to take part in the identifiable as worthy of evacuation in a one- planning than it was to sit around and complain week notice situation His plan was to order a about the lack of it sufficient number of C130 cargo planes to the NSA has historically been an action-orisite to accomplish the evacuation in an orderly fashion He stated that each item identified on ented organization We have all prospered individually and as an organization under this systhe inventory as worth over $1 000 would be tem but times are changing and we need to brought out in a one-week notice situation and he wanted disposition instructions for this change--like it or not The PMD system appears to meet that need Who knows we might just equipment listed with the inventory He indicated that the instructions need not be Return throw in a little MBO to NSAW since the same equipment might be in demand at other overseas sites being augmented to assume the missions from the phased-out site The axiom PZanning is Managing began to surface in my consciousness A PARTISAN PEEK AT HISTORY The discussion then turned to the PMD which was due to be approved in approximately six weeks The Program Manager stated that he wanted the M input to include time-phased actions as required in the recent revision of NSA Directive 25-3 We began to discuss the implications of a phasing-out mode the effect of reducing the amount of operations space as systems are taken off line and the impact of this on guards and alarm systems It was becoming apparent that there was going to be considerable M involvement On the subject of personnel the Program Manager indicated that ten NSA civilians are currently assigned to the station in addition to the contractor personnel It struck me that some of those people may be due for rotation back to NSAW prior to June 1976 If so will they be replaced Will they be extend d If they are replaced where will the replacements be sent in June 1976 Then the thoughts came what about militarizing those billets M2 Or using contractors M52 May 75 CRYPTOLOG Page 6 P L II 86-36 DOCID 4009720 TOP SECRET Ul 'IBRA ARE WE WASTING LINGUISTIC TIME MARY ROBERTA IRWIN G52 2 Stamping the security classification at the top and bottom of each sheet of translation paper Either stamping or writing this classification on in pencil requires a surprising amount of time The sheets could surely be preprinted They have been in the past and no convincing reason for the change has ever been offered the linguist The EXPERT Sheet referred to in this artiale is used only in G but a similar purpose will be served in A by the new Produat Sourae Reaord and in the field by the Sourae Contribution Reaord B Group is aonsidering a version of its own There have recently been a number of articles in the Teahniaal Journal CRYPTOLOG and other Agency publications on language and language processing but none of these has touched upon the problem of conserving the short supply of linguists for purely linguistic work This problem can easily be overlooked or minimized by everyone but the linguist involved with current production on a daily basis Writing as a linguist with thirty years of experience spent almost entirel in rocessin I want to present this problem as I h-a-v-e--s-e-en it and to suggest some possible solutions What non-linguistic work is required of the translator 3 Providing data for retrieval of information valuable for you and necessary to the Agency which takes only a few minutes for each translation This has taken various forms over the last thirty years but with the development of computers the trend has been steadily toward adding just a few minutes more In 1945 the translator attached to each finished translation a 3 x 5 card giving lane message number system and worksheet number This was soon returned with the serial number assigned by some now-forgotten ancestor of NSOC which a clerical assistant entered in the section's permanent records Today thanks to the aii-powerful computer which can save untold numbers of manhours we in Ghave arrived at the EXPERT Sheet simplified to the extent that the translatoranalyst has to fill in only 19 or in some cases 20 of the 42 blocks provided plus the security c JassificatioIl whichm'Ustbehandprinted at top and bottom See Appendix EO 1 4 c f ormat10n P L 86-36 Of all this 1n on 1 y t h e t1tle intercept data system and worksheet number really have to be provided by the translator The security classification could certainly be preprinted and each section provided with those classifications it needed Precedence could be assigned by an 1'5 analyst who in any case often changes that assigned by the translator Other blocks except for two discussed in the next paragraph could be filled in by a competent clerical worker hired for considerably less than the salary of any professional linguist This presupposes enough clerical help to allow the designated person to complete the EXPERT Sheet as soon as the translation is received and to carry it immediately to the 05--another only a few minutes task which frequently devolves upon the linguist It is not permissible to wait until several translations are ready and deliver them simultaneously for work must not be piled on the Flex Room late in the day Typists' time is precious 1 Logging and filing This i of course among the other duties as required supposedly performed only infrequently by linguists when an area is temporarily short of clerical help I would estimate conservatively that this situation has existed off and on for at least half of my thirty years Let me make it clear that I am not criticizing lower and middle management Managers at these levels have always been acutely aware of the problem they simply cannot get sufficient clerical help But translators cannot operate unless and until messages have been properly recorded and filed so that references can be found consequently this work devolves upon qualified linguists often the The two blocks which could not be left to younger ones whose linguistic development unde a clerk are numbers 13 now reserved for TAG's the guidance of older and more experienced lin- Topic and Area Guides and 37 KIQ's Key Inguists is thus slowed down telligence uestions--to be filled in only on May 75 CRYPTOLOG Page 7 TOP 8 8CR 8T UMBRA DOCID 4009720 TOP SBCRBT UMBRA the last three carbons Before the addition of the KIQ's and TAG's the sheet required approximately five minutes Now those two items require more time than all the rest of the EXPERT Sheet since a great deal of consideration is needed for each message What is to be done about these 1 Admittedly translators must be intelligence analysts for they have to decide what is significant enough to be translated and in what order of priority and in addition they must apply their judgment as analysts to the correction and translation of corrupt or telegraphic text But the assignment of KIQ' s and TAG's is a task of intelligence analysis which does not require linguistic ability Qualified SRA's may be as hard to find as qualified linguists but since in any case they must review the KIQ's and TAG's assigned by the linguist and since-and this is far more imoortant-- these codings In conclus on then it seems that we are 1 Stunting the development of our younger linguists by using them as clerical help thus making it impossible for them to acquireon-thejob training under the guidance of thei more experienced colleagues--something which they all need however excellent their college records may have been and whatever they may have learned in NSA training classes As the older linguists retire in ever-increasing numbers this problem is assuming the proportions of an emergency 2 Taking all linguists both young and old out of the linguistic field and into aspects of the job that could be better done by intelligence analysts who do not require linguistic ability In my own work as a translatorchecker I estimate that 1 spend about four hours per week on non-linguistic functions Some translators spend more a few less The amount of time seelns small but it represents 10 percent of my working hours per week If this figure is anywhere near average Can the Agency afford to devote one ofevery tenof its linguists engaged in current production to nonlinguistic operations May 75 CRYPTOLOG Page 8 tfOP SECREtf UMBRA EO 1 4 c P L 86-36 DOCID 4009720 L TOP StJCRtJT U IBRA 86-36 Appendix The Cryptolinguistic Association will hold its 10th annual banquet and awards presentation on Thursday May 29 at the Troj an Horse Restaurant in Silver Spring The speaker will be our own Jack Gurin whose talk is entitled ''What's a Nice Guy Like Me Doing in a Job Like This The winners of the CIA Essay Contest will be announced and also the recipient of the Jaffe Award which is given each year to SOIDe9ne who has made an outstanding contribution in the language field The gala evening will begin with a cash bar at 6 30 followed by a roast beef dinner at 7 30 then the program at 8 30 and at 9 45 a social period which will include dancing All this for the modest cost of $7 50 to members and $8 00 to guests I For reservations contact 5471 s -U-NC-LA-S-S-I-FI-E J D ANSWER TO A reader has chided us for not explaining the usual rules of the game According to the Washington Star these are 1 Words must be of four or more letters 2 Words which acquire four letters by the addition of s such as bat cats are not used 3 Only one form of a word is used 4 Proper names are not used According to our puzzle expert usually an abridged dictionary is specified beforehand O K The following words are to be found in Webster's Seventh New Collegiate Dictionary t Clog cloot clop clot cloy color colt cool coop co-op co-opt coot copy crop crypt crypto glory gory grot gyro loco logy loop loot lory orgy orlop plot ploy L -----------------------'I poco pogo pogy polo poly polycot pool lOP ilil Rii r l ltlllR poor porgy port portly prog root ropy rota tool tory troop troy typo tyro UNCLASSIFIED May 75 CRYPTOLOG Page 9 TOP ECR ET YMBRA L 86-36 DOCID 4009720 TOP 8I3ClRB Ui'ttBRA -----_ _--------------- - ANSWERS TO CLASSIFICATION PROBLEMS o o I TEI ATlO AL AFFAIRS I STITUTE t Members of the International Affairs Institute recently elected the following new officers President-Elect Robert E Drake Members-at-Large I I Barbara W Clark I I Robert J McDermott Also at the March meeting of the Board of Directors a new secretar' Robert J McDermott was elected to complete I time in office The President I I and the Imembers of the Board are currently working on a series of projects aimed at expanding the activities of the LA L to more directly involve and benefit all the members of the Institute The members will also be encouraged to xpress their ideas and views in a questionnairc which will be distributed in the near future 'p L 8 E- 3 6 UNCLASSIFIED I hope this exercise clearly demonstrates the problems and difficu1 ties involved in determining classifications and specifically what information needs Codeword or COMINT Channels protection There are eight Classification Advisory Officers at NSA They are willing to offer assistance in your classification problems but the responsibility for proper classification belongs to each of us My advice then is to use the Classification Manual and USSID 3 consult your Classification Advisory Officer when necessary and most important take the time to determine the appropriate classification and thus to insure that the proper protection is being afforded TQP SilGIU T J tllR A i EO 1 4 c EO 1 4 d P L 86-36 May 75 CRYPTOLOG Page 10 TOP 8SClR13T YMBRA DOCID 4009720 SECRE'f LETTERS ToTBE EDITOR L 86-36 To the Editor CRYPTOLOG I hope you can find room in your next issue for clarification of the edito r i a l__- introductory note for recent p ece on e '--u-se-s-o f -'ELINT April 1975 Having established that in the beginning there was COMINT your note then refers to ELINT as the other half of SIGINT Unless PI has developed a new-new-math which permits more than two halves in a whole you have unintentionally I trust eliminated the other third of our SIGINT family TELINT Confusion on the place of TELINT Telemetry SIGINT is as uncer tandable as it is widespread For years NSAers thought of it if at all as a sub-set of ELINT largely because it used no COMINT coverwords and because in 1960 USIB had decided to classify it and handle it like ELINT A former Chief of PI was a leader in opposing this ELINT-by-association theme since telemetry's job is to communicate information it cannot be a non-communications transmission in terms of the accepted definition of ELINT I On the assumption that CRYPTOLOG is happy to support its sister publications I modestly recommend to your readers a two-part SPECTRUM article giving the background of NSA's involvement in telemetry Vol l No 3 and Vol 2 No 1 Talomatry and How It Grew The term telemetry is often used with a broad meaning whic h includes some but definitely not all other kinds of Foreign Instrumentation Signals FIS a term which is appearing more frequently in USIB community papers In fact there is talk that a new NSCID may divide SIGINT into COMINT ELINT and FIS Telemetry has come a long way in the 20 odd yearsl timely I lintelligence support to U S Navy and other tactical commanders 11 - lv3 The EcUtolL lLeglLe t6uU y ac k nowf edge6 the VrJtOlL and plLOmi6e6 U -6ha U not happe n aga in o EO 1 4 c P L 86-36 May 75 CRYPTOLOG Page 11 SECRET IWIBcE VIA 68P1INT 61b INH6 8JI15Y P L 86-36 P L 86-36 o To the Edi tor CRYPTOLOG I Ile ter in the April 1975 CRYPTOLOG concerning the professiona1ization problems of bookbreakers and the related questions posed byl Jthe Cryptanalysis Edi tor in her note tend to confirm a suspicion that has been growing in my mind for the last five years the venerable term bookbreaking and the importance of the SIGINT produced by its practitioners seem to have become lost in the bureaucratic red tape somewhere in the fogshrouded ivory towers of operational and personnel management This loss of understanding by management goes far beyond the mere fading of a useful term application of the specialized skill itself has become clouded in management's mind to the point where in B Group at least i t i s _ now c10se1v equated with crvotana1vsis I o i j 1I The current CA Task Inventory revalidating the COSC job codes provides an exce11entopportunity to review and perhaps redefine the term bookbreaker and what he does or should be capable of doing in terms of job skills And this letter is my opportunity to offer some personal views towards that objective based on 31 years of experience in almost every aspect of the SIGINT business I EO 1 4 c P L 86-36 May 75 CRYPTOLOG Page 12 COHFIQ8HTIA WNIQbE VIA EURlQUHll' EURlIWI llibe Q lbJPY EDl L1 c DOCID P'Y40093'720 CONFIDHN'fIAJ - I And without the bookbreaker I s alchemy there is damned little substance other than TAl from which the SRA can construct a story and garner the kudos that so often rightfully belong to the bookbreaker industriously working in the background to perfect the code so that even more and better secrets can be revealed The bookbreaker is the individual who created something from nothing tt L - P L 86-36 To the Editor CRYPTOLOG There is need to recognize bookbreaking as a distinct form of analysis by the establishment either of a separate career field or of an option within a separate cryptolinguistic career field The cryptolinguistic option of the present language PQE does not test directly for knowledge of analytic principles use of machine support of predecessors as models of statistics of working aids etc all of which are necessary for bookbreaking And now more than ever these techniques are indispensable for bringing a code to readability for we aM UGall c with lower volumes and at the same time pnP 86-36 difficult codes as they have been improved through the years In sunnning up my view of the bookbre aker is that he or she is first of all a linguist but one who has progressed beyond routine trc mslation It is a highly personal and individual capability Through some quirk of nature or personality he or she is usually intensely in teres ted in solving the mystery of the unknown creating knowledge out of meaningless code groups It is this spark of genius that is the bookbreaker's true value GQNYIQeN rU b II GGQ May 75 CRYPTOLOG Page 13 PI-AP 75-S3-23605 P L 86-36 CON'FIQ N'TI1 1 n lI3bE VIA EQPlHl'F EIWlNEbS SIUN ---- -- ---- --------- DOCID 4009720 TOP SECRET e D CJ '- TillS 90CYMK T CO TAl ODiWORD MATFRuI 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