'FOP SECRET National Security Agency Fort George G Meade Maryland •• -t - _ - • •____ _ _ -·- -- -- ---- - - - - -JI L _ - - - - 4 'P · I Declassified and Approved for Release by NSA on 10-06-2020 pursuant to E O 13526 MOR 107694 Thia is Draqon Seeds There is fantasy irony and the bite of reality in the It speaks of the Bast And like the Bast it suggests lllUCh says little name Dragon seeds is both Mother China and her neighbors Dragon seeds is monumental a nd minuacule It ie the p et · and future It begs for elaboration bl it gives none In it are echoed softly slurred Mandarin brittle Vietnamese determined Korean In it is the spectra loODling ovar the Tha i Lao and Khmer It is frightening and friendly It is uncertain Above all Dragon Seeds is premise It is fert Lle with ideaa unbounded to be cultivated with creativity and i11a9ination It ia challenge It is a live It will be more than it is Dragon Seeds is yours May it grow with you The Bditora • '118P BetJRlff HMBRA _ _ __ J PL 86-36 50 USC 3605 1 SBBDS Publisher DONALD a Blcecutive Bdi tor Managing Bditor Minnie o MC COWN CHI BF 803 Roberts Benj in McNeal Compoait 1 on Helen Ferrone Copy sditor Thaaas L Glenn Biographical Bdi or Rewzf te Bd l tor Geraldine J Pettie $9Cial Interest Editor Bducation editor Brook• H Kandy Marian L Reed• lfe ture Bditor Ray F Lynch Richard v • curiin PRBSS CORPS Bll C rolyn Y Brown 1142 Velma Ja tteracn 812 Philip J Gallagher 843 Mary Ann aalo 821 Gary Stone 844 B31 Jack Spencer M5 John Thomae M Beall BS Paul M r g Joe T • Hadaon 862 833 Louie Ambroaia 863 Jean c smith 834 Tbamaa L Wood 864 Allan Le Gilbert w 865 Leona B Dickey James sahmidt a George U arek s Patter90n 'ft P WltB'f UMBlb TABLE ·•F CINTENTS Captain Joalin'a salutation 1 Captain Joelin'a Biography 2 cryptanalyaia Through PUnctional Linguistic••••••• Donald Lenahan 3 a Stoffel s Impact of AltDP' on Tra f ic Analyai••••••••••••••••••••• Al Gilbert 7 the AG-22 and Yau • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • Peggy Barnhill 9 OP - Decllpe Delete and Progre•••••••••••••••••••• Charles s L t 12 Solution to a Dilemma •• •••• •• L st Clair Myers 14 The Creative 'tranalator ••••••••••••••••••••••••••••••••• Tom Glenn 16 Recovery ot a Viet Cmmmniat Callaign SY•t- ••••• wayna Olin••• Voice Analyzation ot Data ••••••••••••••••••••••••••••••••••Dick Olrtin 91edlinga Ask the 19 22 Dragan Lady 24 ContritN ora 29 • • ··-· ··--- iREETINGS FROM CAPTAIN JOSLIN It I• a pleuare to latroclaae lo lhe B Group reailerAlp u lal'onaal pabllculoa wll« la be tafon111Uve lldfllullve dlverelfted - ' even lalflre81lallf A• 1'1iave already lltaled I• i• y taiUal ouao_ • t of IJrapn Seed to B Group _ nel I • hlply HthdluUo UOllt du profeot ud t • IIClllho•t ret to dae ladlvlduaJ Llkewlae I 'View ll u • eJtCellea opportunity for me to become uqalated wllh you - to lmow wbt yo• esre thl akla r lo • ee what tecfaalqau you explorlq lo olttala • belier reel for wlia profeuloul p o t oaa ont you ad tbo -oaad 'Ga to aee wlaat lllltlatlves yoa pfOIIIINII 10 oapalale of Barri•• CNated by the sl• of B Group dlvenlly of lat - f01B1aJlty of report• ••lafr 1111d the p oleotloa of Dm' lafOl'llallon bave preveated ae fNa o-• u11111loatl11B - • II ' •d f• II ' may ooaoept• ud leC lllllqaee wlalcll profeu oully exoUlq ad aNful to bow Now I• yoar ehuoe for tdv• aad take - to preNal yoar tlloupte • ad ldeu to idve other tile beuftt or your partlcal• ex -tlee aad eaperlonoe or to have yoar 1111e• aloa• u • wered by tlae • • lhpa Lady ead to flad oal wbl ot _ tlal• IIIQ8 I •-ly eadorae tlat• uw 8 Gmap veatare 1111d ea _ '4'1111' fall parllolpatlon la • aklag Drapa Seed8 • provoca lve IINfal ad eajoyable pabllc1111on 1 '118P 8H0BIJf m 111 CA PTAIN HAROL'O E JOSLIN CHIEF 8 Captain Ha otd E Jo tin USN b t i n96 to 8 G4oup tht btnt6lt 06 hu 54 Vt« 6 ins a ftWllbt inctud- o ke y 466 ign• AU with NSA 4ftd tht Nav t Stccur itf G au p W46 o Mitlt4 fl t vict Having tnt Uted in tht N4ur in 1951 ht au ving on Gwut 46 4 Se und C ta OtctmbtJt 1941 whtn he Wa 6 wl t h«d Lawn out on the Pe tty OU ice It -ill cAptu td bfl the Japanuc Hi la 4t plane to te aut Gu a11t t A te apend Lng 45 • oAth a cu« PO ht e uAntd to tht U S in 1945 «dvanctd th oush tht utt4 to Chit6 Pttt r 044lct ctttd 14146 co11talu io1ttd cU an _EJ 4 ign USN in 1946 C4ptain Jo un a qu 4Ll ted Intt p tte T aMtato having 9 adue1ted 4Ao• tht RU4 aian «n9 9t chaot 4-t Anaco tla Capt4' n Jo6Un'4 c« ttA h46 bttft highU9hte d by 4 umbtJt o 4 lli9ftll Lunt 466 l lffllle At 6 6 uch 46 Otputy o tlat Co•bintd Navat Pa ty 4t GCHQ CaMMAAcU Kg o icu NSG Ac Uu tt11 Edze U Sca U u1ch Oe pu l fl O i 1 c to Na ua t Se cWtiti G oup 'Pac l c1 A u i su 1t o it e l to t a-t Spe c c AL Ope ut £011 N4u4L Se cu ity G oup Co M4 d and Otput Chlt 04 8 G oup 2 'IOP 8HQBHrf UMBlb • • • EO 3 3b 3 PL 86-36 50 USC 3605 CRYPTANALYSIS 'l 'HRa JGI P1JNC'l'ImAL LINGUISTICS by Donald P • Lenahan 1 8222 3 WP 8BORIR' lJMRBitk 'P9P SBORIR' lJMBR '_ EP0 _3_ 3-b -3 _ __ 86-36 5 0 USC 3605 It's no disgrace to be a slave It's a disgrace to work voluntarily for aazaeone else ••••Cambodian Proverb 4 EO 3 3b 3 PL 86-36 50 USC 3605 RBCCJIIRRY OP' A Vt BTNAMBSK COMMUNIST CALLSIQI SYSTBM by 5 wa vne Stoffel 803 '118P 8BEJRt1f IJIHIDA EO 3 3b 3 PL 86 - 36 5 0 USC 3605 • 6 '118P 8R8Rlff IJMBM - '118P BfiORtJ'f UMBlb TH8 IMPACT OIi' A RDP' 14 TRAP'll'IC ANALYSIS by Allen L Gilbert 86403 47878 l'be Vietnameae conflict and techniques for collection of signal intall igence developed and employed in that arena have intlwmced the tratfic analytic approach to the Vietnamese Communist problem protounc Uy one ot the moat ef'f'active teclmiques employed on a lar9e scale in Vietnam ha• been Airborne Radio Direction P' lnding ARDP' • ARDP' in addition to revolutionizing the direct support ot tactical uni ta through timely and accurate locating of enemy units has almost reversed the traf£ic analytic approach to maintaining continuity and developing new targets in some a raas aditionally the tratfic analyst is faced with the problem of reconstructing a communications ooaplax through recovery of callsign a nd frequency systems aaage externals schamle activity and those rare compz amises 1111da by enemy communicators Thia route usually requires close ecnt lny and cataloging or the el8Jl8nts of intercept through an extended period of time with the hope that a tranamitter looati011 will be compromiaed or that medium-range direction t'indinc will suggest a location for the activity ARDF provides a location within a radiua of hundrede ot -tare rather than a number o mil••• l he availab l 11ty ot ARDP' en tuget transmitters conaider bly shorten• the period of development or n- activities and provides ai »at instant continuity on targets effecting oonmn1nications changes• In Vietnam the ccncept o ARDP' taaking provides cover• • in all areaa of hostile troop activity rbe Military Assietanae Command Vietnam MACY controls the taaking of direction finding aircraft and has divided the target area into smaller areaa of known enemy activity aa reflected by a 11 intelligence aources Aircr t are deployed to tbe • e areas in support of MACV intelligence eouxces Aircratt are deployed to these areas in suppart oL MA CV intelligence requiramenta and therefore direction finding locations are available allloat daily on tactical targets In this proceaa • a certain numbe r ot unidentified trann l ttera are also located It la apparent that repeated t'ixing of an 7 • unidantl fied trana1itter in the same location even though the callsigns may change periodically gives a ba sia tor a suapected continuity as well as a hint as to the unidentified transmitter in an area where continuity has been lost sug eats that the unidentified target represents the lost continuity When aircraft are daployed to a target area on a daily basis the recovery of the signal environment in the area builds rapidly Certainly all other •le m ents of traffic analysis must then come into play to establish case notations and identifications and ARDF alone does not solve tha problem but what an advantageous beginning it providasl •••• People w«tUKg «toK9 th h4tt6 Peopt te«-ins o the w«tu Peopl ta949ed in conveu4 lon 0 «c Uvt in c tu A lo tc tatian lth «tt thl6 4cUon and • Ulng mob Vou wondt who i6 on th job 8 THK AG-22 AND YOO BY Peggy Barnhill 842 A naw item of the SIGINT community• stocl pile of electronic gactoeta is now installed in su tt'icient quantity to 9rea tly aftect the tra ££ic analytic and processing procedures employed at forward field intercept and within B Group The gadget properly referred to aa the AN GGC-15 but more commonly called the AG-22 replaces the typewriters or mills previously used by manual Morse intercept operators and radl otelephona transcribers It consists of an electric typewriter with a modified keyboard a paper tape punch and in some cases a paper tape reader all connected to a solid state station clock tba AN GSQ-53 The installation of tha AG-22 undoubteclly represents only the first of many revolutionary techniques being developed to permit the rapid trans er of intercepted data from overseas sites to a central processing center The Improved AG-22 Terainal System IA'tS is already being teated at USM-1 Vint Hill Parm Station N Lrrenton Virginia Aa each technological advance is made• changes in traffic handlin9 or processing procedures will occar In order to f'ully understand the impact ot the AG-22 it is necessary to examine the equipment and processing developed here at NSA The AG-22 prodllcea two outputs When the operator strikes a key a character is printed on a pac e and simultaneously the correapmdir19 ccn tiguraticn in eight-level coda is punched on a paper tape Thus total intercept is immediately prepared tor transmission The paper tapes are transmitted via the STRAWHAT data links There are currently six circuits between NSA and intercept sites in the Far Bast then circuits are capable ot forwarding data at a rate of 750 and 1500 words per minute 9 'f8P ffflelffl'f ffltlltltA 'IQP g J RKI' lJMBBt EO 3 3b 3 PL 86-36 50 use 3605 • Thu• the deve1opment of the AG-22 ha• resuLt CS in the ava 1 1- ability to the analyst of total t ratfic only a _te 'hours after intercept The data explo•ion is conaidera ble1 'During testing on various CHICO problems 4 to 5 '5 ti1111116 as 11UC ta were onrardad and processed as would be e q fected 'noJ ID lly via STRUM or Bl FAIR To make both ccmputer an4•analyst pfo easing efficient it was necessary to emp oy compqter eehniques previously thought impractical or si ly i mposeibie · All intercept copied on an°AG-22 o pre ar d in an AG-22 CClllp tible format ia processed tbrough the qltneralized AG-22 Processing Syata GAP whicl ls a series of ' IBM' 360 computer program a The GAP system ••ta ndardizes- codil l'c tantifiea record types asei911ed a processing tri9rapl base4' upon caae notation 0 and provide • various cct erage accou ting a tt lity control listings Baaed upon processing trigraph i GAP data are directed to variou• subroutine programs At prese thlire are five user routine• operatio L i £or B Group problem Processing Trigaph • 8n ti1 ias OU tput Formats QO-' f SLP'AIR BLFAIR • AU otHKR VC FF STRUM TA MSB SBATS I FF StRUM The outputs generated by the user routines are compatible with the existing manually prepared vehicles but may differ slightly in format Data or B Group entities other than those listed above are directed to LBPTOVBR lists which presents traffic in chronological order as copied Bach of the user routines ha• follOlf-on progr81DS which are run prior to the presentation of the data tor the traffic analyst usually le •• than 24 hours after intercept These programs do 10 r I---------- I EPOL 3 • 3b 3 86-36 50 USC 3605 p 8BORIR' HMBltlt much of the preli n t ta ROrting nd listing required by analysts Within the routine which has the la rgeet number or users for examP e 1 that ca se Callaignf a re paged and indicated if predicted for callai911s are matched gainat callsic na from all files processed dUring he last five days The number 0£ files in which a callaign appeared on the same case will be indicated as well as the number ol tiles in which that callsign appeared on a different e se The differing ca sa will be indicated 2 L 3 Files notated l re matched against all cased and uncaaed data tor t h e i t fiVU i iys and if possible reidantified based on callaign usage If the match is against uncased data rbitrary Case Notations AOl's may be assigned based on callsign page usage or two-day continuity 4 P'requencie• 1t nd schedules are pre • ented in link increments and the re son for each contact break is entered 5 Preubles are formatted and in some cases traffic type indicators are inserted 6 Chatter lines a re protiled and 1111ighted to indicate eigni ficance 1 Message address information PAG'e BSD'• etc is isolated and presented in a formatted record a Spacial records indicatin call-up order in multiple call-up are generated All this is dona because analyats and progxammers oot together and let their laa inationa run away with them The limits ot the computer'9 ability to perform preliminary analysis has certainly not been reached As we continue to work with the AG-22 and it • output even 110re capabilities will be defined rhaps some day we may even• • • • Complete this paragraph in 50 words o r leas and submit yow dreams to Peggy Ba mhill B42 11 rDP - DBDUPB DBLBTB AND P'ROGRBSS by Charles w SWift 86 °4 A basic problem that touches every individual in B Group and probably throughout all roduction eluients is the inability to take actions beCAu• e ot missing and misrouted JDeeaaoea TBCHINS 1043 and 1044 provide apecitic 9Uid nce or the routing of information and use of Delivery Distribution Indicator• DDia which should aaaure the proper now of material but in practice many problems aria• due to a variety o causes Instead of seeking out the causes for miasino mesaages many elements have arbitrarily a ddad Il Is on the theory that it their DDI ia on a massage they are assured of receipt Since many DDia ha w multiple addXessees this method actually com- pounds the problem and clogs the machinery designed to provide tiaely and efficient service to N- elements A recent survey in one a Group o tfice revealed that excess copies of DDP material were being received one third ot the maaaa9e copies received were tossed away before they reached branch level Some field sta tiona to rwarded technical support maHages using DDI ccmbinatione that dumped aa many as thirtyfive copies of the message into the ott'ice At least fifteen copies were teased Ula' • and only ive were really required Three factors contributed to this situation 1 P'ailure by the field atationa to select IJ Is according to TBCHINS 1043 and lOM In some instance • the field station had obvioualy ohoaen to use multiple DDI• to assure delivery in other caaea the el'roneoue use had been directed by elementa within the office 2 Failure to assign qualified and dedicated personnel to distribution functions ntatribution was usually treated aa a eecondary mty in most eleaenta 3 P'ailure to provide knowledgeable individuals as the focal point or all meaaage distribution problema to assure that distribution peraonnel at all levels were advised of requirements 12 Jo Since auatez l t i • the word o't the day it would only • - logical that now le th• time to isolate costly trouble apot• in every facet ot operation•• Wain B Group are in a position to do our pa rt now in respect to the handling ot incoming maa• agea We have started in the right direction by providing for one kn01Jled9eabla individual in each office l n B to act aa th• coordinator and authority on all DDI problltlD8 1 both in-house and tield-related Other actions which should be taken are 1 Th• aaa l 91'lmant ot qualified indivicmala to pertorm distribution functions at all levels 2 The constant review and monitoring of DDI requirements by l nclivimal el nta to ditterent l ate· between what i9 required for job pertormance and what I a just nice to have The IDI coordinator and the office of primary intereat would then be into anad of miauaea 0£ DDia and any instances where diatribUtion of material could be remced or elildnatad By implemant l ng thaae procedures a great number ot the meaaage copiea could be eliminated Thia would allow distribution personnel to concentrate more on accurate distribution thereby probably decreaain9 retransmission requests A cooperative eitort by all element• would relieve the preaaure placed on our limited teletype diatriblltion system thus assisting in the tim ely receipt and handling ot our correapondance · Though the hen ma cackle all day she can lay but one •••• ram t h e • • 1 1 - h EO 3 3b 3 PL 86-36 50 USC 3605 CHINBSB VOICB SOLlJrTO T0 A DILIMCA • • • • • • • by L St• 01-a•ir Mye rs • 644 1 Let 1 s take a close tQPk•at NSA's prob a in bopin9 • 0 Chinese voice intercert• -· one ot the lea understopd a n d _problems in the A •ncy S1 oken Olinese i 9 without doubt moi e difficult to grasp t'han the Wl'i tt en language- Natiw speakers are generally the on¼y· ones who pan eully percelve the tonal differences and understand the subtle ••colt' oquialisms •• • • inherent in the l anlJUage However• the uM of pati9e speakers • is not ipso facto a total solution to the proble•i only rarely is it possible to find who can put ha iafonaation down on paper in good nglish c hi• usual procadui e ls to trans scribe what he hears into tbe Chinese charactirs of his native language usually in the alrsive script whiC'h isl a · shorthand • form referred to a a grasll Wl 'iting But uniterstaniling this fdrm I one requires a well-trainey•1inguist -- i - I NSA and tha Service cryptologic Agencies SCAI rely upon of military 188n tr ined in •poken Olinese who interpret what they hear and put on paper in Bnglish 1 their translation of what is transmitted In NSA these arv eJ roneously called transcriptions • Intercept tapes that the ailitll ry linguists are unable to translate muat be sent to NSA tor translation Seldom a re field-translated facts 11 called into question and the only way that NSA can check their accuracy is to request the original tape from the intercept site -- iL it has not been eraaed after the lapse oC the 60 days permitted by current instructions Few other SIGINT probleaa accept the risk of erroneous field translations so truat lly the use I Few of tt se young 111il itary voice transcribers have worlcad with the language long enough to devel op the voc bulary or experience to cope with colloquial words or phraaea that go beyond the routine stereotyped militaxy language for which they have been trained Furthermore neither NSA nor tba Service cryptol0c 1lc Agencies SCA' are likely to expend the time and money required to develop the large number of really expert linguists that are needed at intercept sites to translate tranaacribe voice intercept with the degree of accuracy that NSA 1 s mission requims 14 - EO 3 3b 3 PL 86-36 50 How than can the pre•ent syatem be improved • °The optiJIIWD is to have thor0u9hly seasoned linguiata at the potn •of btercept lacking thia capability the system lllllat lM ecl to the capabilities ot tbe personnel involved A• no tMtd f'bove ·P rofessiona l linguists at the point of intercep •are•not lilcely to be provided so it may be neceRSary to contider plan utilizing lees qualified peraonnel and change to - - t clmique Y using PINYIN las a ima-eaving•short- a hand device where posslble mllltary linguis •tran• criber in out true transcription that retain -the original terminology tor analysts to check •-- if chec1Q ng is the field could turn • really necessary • PINYIN is the official CHI X»I roaanized spelling- of sounds in the Chinese language tor a variet of raa90na both PINYtN - • invanted tn 1957 a c lcal and lltica 1 • __ _ __ • • Translation o tile f lNYlN '---- -------- is the next problem to be conaidarect The rs coa tranalate the material ae t bey go o» • later in the J 'ight-hand maxgin or directly beneath tbe Chin••• PINYIN In my opinion written translation• are not •necessary until they beco •••ential for the analyst or raPC rter's understanding of the transmission or tor inclusion in a- SIGINT product report Visual mental translation ahou1a be sufficient for moat analysts familiar with atereotyped texj - and it ls not as difficult as one might think In e f•ct tra ttic analysts reading Moree and teleprinter chatter r-1a ra right now doing juat that -- reading Chinese even 1J-lii eitectly And if all analysts are thus forced to absorb soma alight knowledge of the Chin••• language in order to do their job and do it better wouldn't this ba an additional benefit to the Agency And who knows how many of the•• non-linguiat analysts might develop into campetent linguists after tonaal training in the l n911 9e 15 • DIE CRBATIVB TRANSLATOR by Tom Glenn 861 It is aeeential that tha oapt in take steps to assure a n attack a a aoon aa poa•ible •• the translation read No delay will be accepted I knew tbat the text in question has been paaaed in the heat of battle by a man daaperate in the ace of imminent de tea t nd possibly death It atruck me that his language was ra tber tormal t01 the occa aion Tha original read DI I J UYS DANHS CHO DWOCJ CANGF SOWMS CANGF HAY I All' CHUR YBBDS CHAAMJ ZIP' KHOONG IM CJ I would have t ranalated it Strike sooneat without t'ail Time is o the ••••nee Any delay will mean failure The fii'• t translation was not wrong · It simply miaaed the point The example ia an extreme one and it has bean somawhat altered to protect the guilty but it is symptomatic of a tendency of translators to ·9JIIOoth out the unruliness of the original to impose order and busineaswlike oalm to exp resa everything in unrut'fled government Bnglish When we do this destroy the vitality o the original dehumanize it and distort it In so doing we do our customers a distinct disservice to say nothing of insulting their maturity This article then i• a plea for more creativity in tran•lation Unlike other diaoiplines where there 1• only one right ans r translation plunges the practitioner into the world ot aabigui ty where there are p l enty ot wrong anawer a and many right ones The choice of the JDOat nearly accurate answer depends not on dictionaries grammars and TBCHINS but on intelligence emotion and undarst ndino For translation is rooted in language which ia first add foremost a sensual thing ir retrieveably tied to feelings in the chest throat mouth no- and ears and heavy w l th aotional cwaa But langua9e is also our primary aeana ot information conamrtication and bringing inc 18 auto darkn•••• And aa any linga latics student will tell you language 1a erratic ayncretic and dynaaic In coping with euch a n animal creativity the ability to deal with the unknOlnl and £ind new anawara -- ia simply necee• ary 16 There is nothing pardcularly revolutionary in a ny ot the foregoing But surprisingly we largely ignore the need for creativity ln tranRlating our rule-ridden proaecmree with programmed solutions tor all eventualities are in pa rt to blame We hav 11ou9ht •• moat organizations do to take the the uncertainty out of our daily work by promulgating proper procedllre appropriate to whatever cirCWDatance P'or translating it is a futile effort P'or no matter what the girth ot our guides gloaeariee grammars dictionariae memoe lexicons manuals and primers we cannot replace inventiveness with rules But since we have tried so valiantly to do so we can hardly blame our translator• for believing that modalite is always translated a • procedures or that_ 4 J fl invaz iably mean• exploit What wa h ve done in ettect is make admirable progress in achieving ma china translation from human beings Portunately it doesn't work One reaaon is that words mean such ditterent thing• to di terent people As an Iriah nu r• e once explained to me the screw f n the Briti•h Isles is slang or wages Similarly ''Defense de trepa aser aa a aic n on a cemete cy gate in C nada announce• means no trespassing not no d3 ing in the same paz t ot the world chars us 9es means used care not shopworn chariots Tbe in fluence ot Americans on the nations ot Soutbeaat Asia hae produ ced new hybrids A sign in Saigon w J na Paa de tumer n'est pends -- a ll'l'ench version vi Vietnameae of the redllndant military Bnglish No smoking allowed In aoae oriental language• it ie i mpolite to answer no to a superior Thus a Vietname11e who worked tor m e in S igon in trying to adapt to American caaualneaa anmred moat ot my qufftiona 1'Da khong a -- Yea no sir In Bngliah no doubt otten means there is some doubt at chance maans t1111 ll likelihood and Su rely you don't mean that means ''My Godl You mean that t Despite tbeaa and other problema tranalatora peraiat in trying to program themselves We could help thea in three ways P'irat • should 911lpbaaiza ma• tery o£ Bnoliah a factor in tranalation we ha ve overlooked with dogc sed consiatency F'b at cate translation atter all requires a p rotound understanding of tbe way oliah works how it can be driven shaped cut and tooled to make it catch the senae and feelinc i ot the orJ ginal 17 • 'f0P B BORIH' UMBM Second we should t%ain tranalators -- and eapeciallv B Group translators -- to become comfo table enough in the foreign language that they can •ever their dependance upon Bn91ish to understand foreign texts In essence translators should be able to read a sentence in tlMt target language understand it wt thout reference to Bnglish and only take up the question Row do we aay tha t ln English Aid• to reaching this stage are a good ear well tuned to the sound of the target language a willingness to grasp at the basic meaning of a word which ha s no equivalent in Inglish la i in Vietnamese for example has only one meaning not the halt' dozen dictionaries give and ability to think without reaourae to words Voila can best be understood in terms of geatw -e and facial expression ''Khoi in terms of picture of a circle and things outside it Third we should enccuraoe eros9 trAining of linguists ideally in related languagea Chinese is the Latin of SDUtheast Asia knowledge of Chinese ia a valuable 11 sset to Vietnamese Korean and Japanese linguists who 111USt struggle with borrc Med words of'ten vary d lt' ficult to translate Thai and La a a re closely related And so on inally and perhaps most important translators mu st learn to unleash their minds Rote translation works tor some tmcts all or tha tille and all textR sOJlle of the time but not for all texts all of the t bae It is at this juncture that creativity the choice ot right phrase or word in English to match the thought and flavor o the original -- becomes crucial A translator hath nedll to lyve a clene lit and be ful davote in preiara and have not hi• wit occupied about wordli thingis that the Holi spiryt the autour oC wisdom and kunnyno and trutlie dreaae him in hie werk and autfre hia not tor to erre ••••••Wyclil lR ANAL'YZATION 01' DATA by Richard V • CUrtin 811 An analyst should tirst atuey data in ita original form looking for obvious or significant paints By all standards it is most important that an analyst look for virtually any and a 11 signs of unusual concli tiona which c011ld occur in any form ln a ny data customarily a thorough analyaia ia a primary goal but prior to any thorough analytic study nmch can follow troa initial scanning of data looking for virtually any important sign or signs Do this first 1 From this point particularly having run out of initial scanning of data n analyst who works with traffic ahould dirty his bands by etually handling and sorting traffic in its original hard copy form Going through traffic occasionally volua l noaa amounts ot traf'fic ls a dlity of all a nalyats Having to do thia baa ita applications to follow-on analysis In this follow-on analyais many sau nd conclusions may solidify by improving facts first found during initialization • 1Llst to avoid con fusion analysis is not sorting traf'fic -- it ia a logical accounting tor all indivicll l parts of a main body of data Kncaing functima and llmits of said ind lvidllal parts is important Looking at all part• individually and aa a group is also moat important Misaing parta could focus on basic primary origins ot data Non-association ot parts could add support to analysis also Odd or unuaual conditions should aid in procmcino a working copy 0£ an original bo f oa which vour data was tor thcaminc Priaarily in addition to analysis ot data an analyst muat list all significant facts for historical background information Quick logical draw back of this in£o mation i• an important point in analyza tion Random approach to draw back of data ia not satisfactory in moat situations 19 Should various arts and skills apply an analyst must vary his attack accordingly thi• is a ai jltl of a good analyst -pliability or adaptability to situations and changing conditions Until an analyst displays this quality in hie analysis an analyst is not tunctioning at a 11l Xi amm standard Vital to all analysis is a thinking analyst with ability to occupy his mind with various and eundry points Which point to disavow or disclaim and which to follow-up is not always obvious X-ray vision would aid any analyst in both scanning of data and looking into goals ot tOJDorrow You as an an lyst occupy• vital position in an analytic community -- much o your analysis is original with no duplication by co-analysts thus your analysis is primary to nalytic community goals and missions Z-groups and A-groups of valid data groups should aid cryptanalysts in locating indicator or discriminant groups and in tum aid in important cryptologic t'indin9s Did you do any analysis of data Bdi tor• s note We will have turther comment on this article in the December issue ot Dragon Seeds PLAIN BNGLISH one ahould hypereathetically exercise macrography upon that aitua which one will eventually tenant if one propels one'• self into the troposphere 20 CRVPTO-ICRAMBLE u _ IICh - to Rk hMJ A _ of 111a ftva numblnd crypllHa'ambla pllclngane letllr' In form five wardl or _ _ of which fill the dlffnldon to a rtt rt B I LLI TEAR Subltltutlon metllad hwolvlng two _ chlr clllil for OM plain chnclll TRI PEBAI T -0 ----0- Systam In which 111• cipla una may be divided Into tM _ pa11 each wltll clwty functions 3 DCHI P oAGRI ________ Subnltutlon mllhad In which the plalntut unftl lrN1lld • pain of daaulliH •• 4 cVAi ___NSTAR c ___ Two or more clpla lYfflboll which _ the plain eq1 IIVllent 2 6 _Q ___ Q ___ ANAID Q ____ RYE PNIIMII wlhldl prodUCII di dlltrlbutlon 111d 11aditk Now 111eclrclldl8ttilrl to form die cryptmnaw• by111eatDanat the right PrlntCRYPTOANSWER lwe 21 8EEILIICS ----The 10th Profea• ional Qualification Bxamination in Cryptanalyaia will be given on Monday and Tuesday 15 and 16 November 1971 PeraoMal who wish to t ica any or all parts of the exami t1on ahould contact the CACP · o f fice Roca 3A116 3868s by 5 Noveaber 1971 Anyone intet• ••t•d in attending prep or tbe examination should con tact Al Verbitz B03 an 5396•• ••••iOAa ----flle Lan911age Career Panel i • investigating the possibility of requiring all candidates for certification to damonstrate their ability to understand their target language as spoken t'ormally The requirement would not be rigidly imposed tor several years 1975 ia being 11antioned to all° for the arrangement o proper training -•-•''The SBATS Message Log ----PQB #Swill be adlDiniatared Bllilding a Cryptanalytic Tool by the Tra tf'ic Analysis career published by 865 is an excellent Panel in tba north aide of tha summary of the current SBA TS NSA FM oat'etaria on Monday proceaaing cycle Although Taeaday and Wltdnaaday 1 6 7 1 and written tor cryptanalyata 8 Dacellber 1971 A new TACP tra1'fia analysts who'd like to atudy outline has been prepared undar• tand more oL the prooeaaing for d htribution to all a epirant8 behind-tbe-•cen•s can learn from To aacertain eligibility candi- this well-written report It's d tea should aubmit PQR' • B65-SSR-02-71 dated 15 Auou•t addenda and reports to the TACJI published by 8654 o fice A 0111 1Cl90 1 by s November 1971 ----A acapilation of varia11• briefing• given dlartng the P'ebrua ry 1970 ''Traf'f'ic Analyais Mechanisation Fc 2 ua was pw liehed recently Twenty-tour briefinta• -re delivered at the 22 Tffli9mttfflJMBltA t'o r wa including fe U% 8 Group presentations r Introcllction to TIA Machantza tion R s Benjamin Southeaet Aaia by Fred Maaon and Dick Alexander and ''Vineland by Dick Wilechka • Copies may be obtained troa Mrs • Gloria Chiles Pl4 58688 • ----Virginia Jenkins Bl3 who ia developing the new course • actic l Diagnosis - CA 260 which deala with the cryptanalyaia o hand aystau and cipher devices and features operaticna 1 problems is aoU cit lng input frc a a GXoup OE particulax- interest are cipher ayate• a aaploying non-cyclic additive• and rail-tence o r grill transposition •s ----The Council of Learned Organisations CLp is planning a eympoaium to aoquaint the NSI community with tha interrelation11hips th- t exist among the major cryptologic di aeiplines with camputer rving aa a unifying tbaae by mean• of lecture• exhiblta and tours Tbe event is achemled to take place in March 1972 ----In an effort to cultivate profeasional linguistic activity throughout tba cryptologic community the crypto-Linguiatic Association is encouraging the formation of Speoial Interest Groupe Por particular• contact Dr Amelia Murdoch 4767s he pilot class in CA•260 bald between 15 March and 17 Nay 1971 The next clue is scbamlad Eor Maroh 1972 • Persona interested in attending or who have subjects for incluaion may contact Virginia Jenkin• on 8-8016•• i I ' 111 ' · •· ••• Articles tor publication may be eutmaitted through Division Preas Corpe maber • or directly to mAGaf SBBDS 803 EO 3 3b 3 PL 86-36 50 use 3605 ASK I THE • • Daar Dragon Lady • • • • •• gRAOON • 1111 L f OV In planning the ttuieraxy for ltorthcoming viait to PBKING Pleue trv·to arrange a n opportunity • for him to visit the 9' te of Heavenly Peace in the old city • • ot PBXIMI ' • _ _ I f t h e t i me ·-be f un d r he • hou 1 d a l a o v i• i t Diring the period of his visit ha might be able to arran9e a aide trip to both PBICHING and BBI JING enroute to or tr _ the airport I niring the fstop at CANTON a· short sightseeing trip around Kt1AJlfGCHDU and GUANGZR U should also prove interesting LAWRBNCB ST• Cl AIR MYBRS 8441 '6378 The Dragon Lady received the following two letters reterring to problems of terminology so ahe paaaed than on tor authoritative comment to our A-number-one gloaaarist the Guru and Caudillo of the lllndee society Lambros Callimahos 24 Daar Dragon Lady What is high-grade traffic Is it that t 'J 8 which i machine enciphered Tf ao why do areas with no machine enciphered traffic categorize some ot their trattic as such Is it the type in which an additive i11 applied to an already enciphered text What if the generation •thod of the additive stream is exploitable ts it user-related fhese a re Just a few oi the ambiguitiea I have encountered The BASIC CRYPl'OLOGIC GLOSSARY June l96S defines high-grade a a Of a cryptoaystam offering relatively great resistance to cryptanalyais How does one measure relatively great resistance ta it in the eye of the beholder If so than all of the above truly be high-grade can you o ter a more precise meaning for this o t used term CAROLYN Y • BROWN B1122 Dear Carolyn In answer to your question let us examine for a DIOJD8nt three definitions aa found in the first 1955 edition of the Bllsic Cryptologic Glossary lo r-q r ada adj Pertaining to a cryptoaystea which offers only alight resistance fen example 1 Play air ciphers 2 single tranapoaition 3 unencipbered one-part coda a medium-grade adj Pertaining to a oryptoaystem which o tfera considerable reeiatanca to cryptanalyaia for example 1 strip ciphers 2 doUbla transposition 3 unenciphered two-pa rt codes ''hi9h-9rada adj Pertaining to a cryptoayatea which offers a maximum resistance to cryptanalysis for examples 1 complex cipher machines 2 one-time ayatema 3 two-part codea enciphered with an additive These da tinitiona were dropped tram the second 1965 edition 25 'f6P BtlORiff UMB A • because obviously the relative security of a cryptosystem is in the eye of the beholder a single transposition cipher may well be a hiqh-grade system as for example a certain German World War II cipher and a one-time system may be low grade if the Fibonacci generating-group even it' randomly chosen ls sent in the clear as the Al group of the message A two-part code enciphered with an additive book may be a high-grade system if the code book is unknown the additive book large the indicator groups enciphered and the cryPtoperiod changed frequently but if the coda book is lmown the additive book small the indicators aent in the clear and the keys in effect for a long period the • ystem would probably qualify as a lowgrade system All or the foregoing rmaarks apply to our cryPtana lysta a cryptanalyst from an emerging Atricanrepublio might ind it impossible to cope with a Play£air cipher and ao as £ar as ha was concerned it would be a high-grade system And s- Carolyn the dafini tiona for low- medium- and high-grade cannot be made more specific since they so subjective e Dear Dr gon V Lady1 I am a nawaomer to the world ot manual cryptosystema and the jargon hae me completely con sed There 1e a def'inite terminology gap between the honorable elders and the neophytes lilca myself who have just completed basic CA course In tact there even seems to be a terminology gap between the different training courses i e CA-100 CA-400 CY-100 • For example what is biliteral substitution The 1965 edition or the aaaic cryptologic Gloaaary defines it aa enciphermnt by substitution methods in which the cipher text units are pairs of cbal actera What about the plaintaxt units It the size of the plaintext unit is luger than one element 111adial plus final or -dial plus t' lnal plus tone ia it not now digrapbic Sllppose variants are eq loyed on a dtgraphic system where plaintext unit size ia larger than one is the ayatea di graphic with variants coda chart with variants or ia it all lamped under bil l teral with variant• 26 119P SBOBR'f IW 7 Is a dinamic syst8'11l included under biliteral or di graphic•••• or both ' And finally -iv has the term uniliteral replaced the term monoalpbabetic which is unlveraally recognized and accepted How does one rater to the basic units of cipher ''lJnilits Bi lits Where ia the naw crypt glossary HBLPll mVID J SHBPARD Hll Dear David Firet of all oavid you msat realize that same honorable elders are just older but not neceaaarily wisex - have eoae first-class technicians who would flunk treahman anc liah Again the terminology gap between different training coureea is a function ot the glosaarial erudition of the particular instructor Now tor your compound question In biliteral aubatitution the cipher elements are pa irs of characters r119 rdlesa of the size of the plaintext elements which •Y be single letters pairs of lettexa or even units of larger size in di graphic sabatitution the plaintaxt elements are indivisible pairs of characters l'egardlesa of tbe size of tha cipher element• wb ich may be tor example pairs o letters trinomaa or other CODtbinations A Pla air cipher is di@aphic becau•• the plaintext element• are indiviaible pairs of latte rs biliteral becauN the cipher elements are pairs of lettera and monoai tabetic because there is a one-to-one correapondance between p n and cipher equivalenta --although the latter should not be at esaed lest it oon a YOLUl9 impreaaionabl e minds or incense older stultified ones 27 • 'IQP SBORJff UMBltA A digraphic ayetem with variants la self-explanatory aa is coda chart with variants if a coda chart were involved ''bilitaral ayste11 with vari l'lts would usually imply variant coordinate• in connection • th some kind ol a cipher square A dlnomic aystem ia either a biliteral ayatem or one in which pl intext dinomas are aubjectad to turtber crypto9rapbic treatment A aimple 1111bstitution cipher is manographic becauaa the plaintext elements are single letters uniliteral because tl w cipher units are ain9le lettus and monoalphabetic tor reason given above Ciptier elements are called characters d 1 orapha trigrapha etc the third 1971 edition of the Basic cn ptoloqic Glossary which bas just been completely revi nd ahould be printed and distributed di iring November Any i'urther que• tiona It is only when there is some mortal deserving ot being delivered that the ain9le live hair of the moat excellent Buddh protrude• itself and stands forth in a straight li tr0111 bet-en the eyebrowa ---the Manual of Buddhiea 28 CO NTR1 BU TORS PEGGY BARNffI U a d4 t4 r 6 tt JU « df 4 t in 842 u a 1966 9Jt Clrfwl te of Ma Jt flwood C0Ue91 SM« n to11 Pa She completed the SR 1nte-tn p-t 09-t 411 i i 1910 «11d l6 ca t-te 11 Ufl woA lt l ns on tlte o coa te 6p1 cl'J c4 tiou fo t the AG-U pJtoce u t1t9 4 f4 t1 11 w t th l 11 8 GJtou p 01 CIC CURT1 N De put Cll le l 811 ente -te d o du bJ w c tla NSA ln 19 50 A t itr i tia tty wo tl in9 Oft the Sovle t p tabte ' he wcu 4ele c te d lo-t the f lu t C l4 44 o CY-100 a nd 6ub 6e que tt tt11 be ca me involved ln c 1r yp ta na l gtlc « tt 4cl OK va t loU 6 A B a d Q p t o te 114 He 1u ce ivtd 4ft NSA 4cllot4Ufl l p to co111pte te la u baclltloJt ' s de 9 r e e ' ° ' i n 111a the 11 Uca t 4 ta U 6 t i C 6 tht Ge o tge IJlul i n9 ton Unlvl t 6-ltf MJt Cu tUn u cM ti l i e d by the CA TA 0 t4 Sg- s te u and M4 thu« tiu cu e u Pcttae t 6 • AL G1L8ERT B640J ca11e t o IISA l n 196' a If tu 11 t UJr £119 l to111 t t A t 11r Se cwr l tf Age 11 cy cu a CIOS Wlt i t e ln ASA he 4e tve d tn Eu topt the Fa t Ecu t SE A 6 la and at NSA roo1L l ltt4 a t v« t toU 6 tbtu 44 1te p0Jt te 1t t a U ic a n tv t Ru 6 i«n t 4 icf i u u cltf pt4 4l f 6 t MIL Git6e t t olto lj pJr olt u l o ze d cu « Sr ec l a t Rue « teh Ana t h44 a o ike d on tlle V-le t11a 111u e Co1111 n l6 t mUJ ta Jt y p1tobte 11 i Kee 1966 o TOM GLENN Oe paty Ch le 861 Ii cu Cl tot4t tlt U tu n f 4 6 e x• pe -'l le11ce w c tla ASA NSA on the 11 Le tllPUt i u te • • Ht U 4 p io4u 1 £ona U ze d 4r e c l«t -te 1ea 1telt a MLf 4' t and V-le tN4111Ut t l NgU U t coho 4l 6 0 6 bl tUf d ChlKU f a lld F-tf KC IL OII lt u OIOII • a Mlf Gttnn ha a uve d ali t•e cltca lu«n ol the 1'le t11«MUe L4n9uge PJto4U 6 0114l lz4 t l Oft E «lfl C 4 tlOM CoMmlttte A44 l 9Ntd to V l etu• J96t-1965 1961-f961 «nd 1969 ht l u bt e n l11vo v1 d ln bulll lc an 4lf14 U c typtoUN9 Uue6 in te t U9enee 4ftUf 6 l6 and mo t -' lsn t l icall-tty - i11 tl e • a K«9ute n t o tlLe S1GlMT Jtt po - t l q eUoJL t OK tilt 1'ie tft411 c 14'1 OONALO LENAHAN a c1tfp tana tv 1 t l K 8ft on the CHZCOM --- b11 obtu ente Jte d oK dutv wltlt NSA in 1961 He eompte te d n t u n pJt ogJta 111 wiat tU 6 £ 9ue n t6 ln A 8 cind G G11 ou p11 4M u p1L 01e u £011• aUztd c1r vpunal 1 4t He l otd 6 a s s i 1 a ' 0111 M4nl « tt4n C0Ue 9e c u1d Ge 01t91 tow11 Uft i vf lU l tf u roo11 l i 11 9 on « 111cutu' 6 de IJie t a t • • EO 3 3b 3 PL 86-36 50 use 3605 8P WRtlT lJMffltxA · 'J9P SHORWf mHJRA ·i----------- EO 3 3b 3 PL 86-36 50 use 3605 L ST CLAIR MYERS Cfl lCOM boalib te «1 u 111d 1 • • CooJ£cUnato-4 l 11 844 i a au·o a qu•t-iUe J1-ooi6Jt e Aie Ji and c11 11 toU•su ut n RUU C 4K cud J4Jj4KUt 1' 119 la £4 U S Naufl 6e Au £ c e • c• •_ dri_ e- a4c _ -_ t e I 19 41 - of 19 u J - o u # CIJle tla t CHI COM l_ o V4 U OIU MSG 11et tv u e 4ft o arl 've d OU c c e • c 11-c u 11 ge 4ftG NSAPACREP • - • · AVNE E STOfFEt B0Jt he g4K ti A C fptoto9 £c tlpt tllCt 4ft 1946 -ltk th ttt •ft41l toull tke A•f Se cullity Age 11cg • At NSA lat • Kollu 4 Wollktd on tilt Sovle t p bblt • un tl t 1954 and on A-1 £411 ta gtt 6 ke J ' e 4 tVt MJt Sto Ud • 11 11e 11be Jt o tile TA C« t e t PaAI 4 t0II 1965 to 1971 4fti k4' bttft 411 46-IOU4tt td toll o COMMANP -llftCt J 961 He l ot d4 a -a S c Le g te e ln pf t 4 lC 4 u111 JokM ffopU 6 Ui lvt t6 l ty 4Kd A Cf• U4 £td ft ScltftCt Ca te e J• Fle tda • tile TA CA SRA l SQ t11t d • Plv a C c«t PL 86-36 50 USC 3605 8BORIR -H MBlb I • • •••••• Jdli la lf rr•u 'JJOP SECRW'