TOP Sf3€RET JtJN 71 National Security Agency Fort George G Meade Maryland - Declassified and Approved for Release by NSA on 10-06-2020 pursuant to E O 13526 MDR 107694 - - -- -- -- f ' This is O agon Seed4 There is fantasy irony and the bite of reality in the It speaks of the East And like the East it suggests much says little name V agon Seed 4 is both Mother China and her neighbors SeecU is monumental and minuscule It is the past and V agon future It begs for elaboration but gives none In it are echoed softly slurred Mandarin brittle Vietnamese determined Korean In it is the spectre looming over the Thai Lao and Khmer It is frightening and friendly It is uncertain Above all O agon Seed4 is promise It is fertile with ideas unbounded to be cultivated with creativity and imagination It is challenge It is alive It will be more than it is • O agon Seed4 is yours May it grow with you The Editors 'fflP elt ET UMBltA I PL 86- 36 50 USC 3605 1 DRAGON SEEDS Publisher DONALD E MCCOWN CHIEF B03 Managing Editor Executive Editor Minnie M Kenny Roberts Benjamin Feature Editor Biographical Editor Richard v Curtin Jane Dunn Rewrite Editor Education Editor Victor Tanner Marian L Reed Special Interest Editor Ray F Lynch composition Helen Ferrone Lorna Selby PRESS CORPS B11 Carolyn Y Brown B42 Peggy _Barnhill B12 Philip J Gallagher B43 Mary Ann Laslo B21 Gary Stone B44 Jack L Thomas B31 Jack Spencer B45 John E uzarek Thomas M Beall BS Paul M Hoagberg B32 Jean Gilligan B62 B33 Louis Ambrosia B63 Allen L Gilbert B34 Thomas L Wood B63 William Eley B41 James w Schmidt ' r · l'OP -wttm tJMMtll 'f0P BBCRB'f UMBRA June 1972 Vol 1 Nr 3 1 i 111LE If Ct TElll I The •If of Staff ' 2 Mayb lli •s Related to the Phase of the Moon The Reality of Communications Changes Herl Guy B45 5 E E Orr B41 13 s Chun B44 17 Dr Ralph W Jollensten Pl 19 Nor _Wild B03 23 Study of ZFK Message Activity Kenneth Miller B43 27 Vietnam@ Communist Tactical COMI f r Uons Tim Murphy B 6 32 A Need for A Centralized Transcription Operation The Open D r The Role of Mathematie9 · • in C A Machine-Aided Translation l See4_1Jiag Ask th Lady Dragon Con 1 utors Richard 34 36 39 CHIEFS OF STAFF E LEIGH SAWYER CHIEF B02 DONALD E MCCOWN DELMAR C • LANG JOHN B CALLAHAN CHIEF BOS CHIEF B04 CHIEF B03 NLike men crossing streams in the winter How cautiousl As if all around there were danger How watchful As if they were guests on every occasion How dignifiedl Like ice just beginning to melt Self-effacing I Like a wood-block untouched by a tool How sincerel Like a valley awaiting a guest How receptive I Like a torrent that rushes along And so turbid --Lao Tzu 1 'JQP SISCRIH UMBRA OP 8BCRB'1 UMBRA E LEIGH SAWYER Ch- -e 6 So 2 --------- EO 3 3b 3 PL 86-36 50 USC 3605 E Le -lgh Sawye 11 ha 6 be en with the Agency ' 6oJt 21 yeaM He wa 6 1te calle d to active mil-l ta Jty duty with A 3A -ln 1951 and c onve Jtte d to civilian 6ta tu 6 two ye a Jt 6 la te Jt In• 1951 he le 6 t h-l 6 cu 6 i gnme nt a Exe c u t-lve 06 6-lc e 11 o 6 the 'V-l1te c Co1t' 6 Plcut 6 a nd Ope 1tat-lon 6 S ta66 and tJtan 6 0e Jt1te d to the A COM Te c hn-lqu e 6 GJtou p In t h i • eapaeHy he wu d i eeUy i nvolv •d cthe p1toble m 0ou nda t-lon 06 the e me 1tg i ng Chine 6e Commun-l 6t Following a two-ye alt tou Jt a 6 NSAPAC O Lnawa Jtom to 1961 he wa 6 de tailed to 6e Jtve a 6 the Age nCJ JSPC P1toje c t 066 i ce Jt a nd a u thoJt-lze d by the V-l11 e c to11 to a t btde pe nde ntly -ln be hal6 o 6 a ti e c he lon 6 a 6 a me an 6 06 acc e le Jta t-lng ac tivat-lon Upon c ompletio n o 6 th-l 6 plt o j e c t -ln late 19 6 2 'he wa 6 a 6 6-lg ne d a 6 Ve pu ty Ch-le 6 06 the 066 i c e 06 Eu 1tope an SAtell-lte 6 In 1965 a t the pe Monat 1tequ e 6t 06 AVP he Wa 6 Jte a 6 6-lgne d a 6 Chie 6 06 B21 Chin u e Co mmu n-i 6 t I and 1t ema -i ne d -i n th-i -6 po 6 -i t-lo n until '1968 S-lnce that tzme h ha 6 6e Jtve d a -6 Ch-i e 6 06 802 i •• rh i ng l MJt Sawye Jt g1ta du ate d 61tom Ha 1t va 1td Un-lve Jt 6-lty -ln 1943 w-lth a B A deg1tee and 6u b 6e qu e ntly 6ollow-lng milita Jty 6e Jtvice in China du Jting Wo1tld WaJt II Jtec e-lved hi 6 M A de g1te e 61tom the Ftetche Jt School 06 Law and V-lplomacy P1tio1t to hi 6 1te ca ll to m-llita Jty 6e Jtv-lc e in 1951 he tau ght h-l 6to1ty gove 1tnment and inte Jtnat-lonal Jte lation 6 001t th1te e ye aJt 6 at the Un-lve 1t 6ity 06 Conne c t-lcut •••• VR VONALO E Mc COWN Ch le 6 BO 3 VJt Mc Cown' 6 SIGINT ca 1t e eA began when a 6 a gJtadu ate 06 the 1n6antJty OCS at Ft Be nn-lng he wa 6 a 6 6-lgne d to AJtt-lngton Hall -ln Se ptembe1t 1942 In 1944 he wa 6 t1tan 66e 1t1te d to the London He a dqu a 1tte M then to PaJt-l 6 and 6-lnatty to Ru 6 6e l 6he -lm VJt Mc Cown 6pe nt the w-lnte Jt 06 1945 46 at Btetchte y and te 6t thi 6 bu 6-lne 4 6 in the 6p1Li ng 06 1946 A6teJt an -lnte1t4egnu m he 4etu Jtne d to NSA -ln Nove mbeJt 1956 pe nd-lng nine yeaJt 6 in AS then 6e ve 1tal a 6 Ch-le 6 84 and moJte 1tec e ntty a 6 Chie 6 803 VJt Mc Cown '4 p1te v-lou 6 caJte e Jt WM a a Ne a4 Ea 6 te Jtn a4che olog-l4t Hi4 6tu dy 06 chemi 6 t1ty at the Un-lve Mity 06 Ca l-l6o4n-la Be Jtke tey WM -lnte JtJtupte d by two ye aJt4 in Pale 6Une 2 'fOP 8BOR B'F U IHR EO 3 3b 3 EO 3 3b 6 PL 86-36 50 USC 3605 in 1929 Hi a11 ted a c heotogy theJte and in TJtan -1011 dan and in 1933 jol ed the 011 ien tal 1n6tltute 06 the UniveJt4ity 06 Chic ago Five qtall 6 we11 e 6pent at Pe11 6epall6 in lJta n and then a PhV Wa4 a hieved ju t be6aJte WaJt ld Wa1r 11 A Guggenheim Fellow hip l n 194 47 p11 ovided a 6a 6c lna ting winteJt in New Velhl the Indu 6 Vcr t ley • Jt a q and 1Jt an VJt Mc Cown then 6pen t two wln teJt4 in I a n combining in 1949 the opening 06 a majoJt expedition t Nippu in 111 a q A6 Vl11 ec t0Jt and a n A66oc iate PJto6e64oJt h continued theJt e until 1954 when he 0inl6hed nece46a Jty publication be6oJt e 11 e tuJtning to the Jte6eaJtc h 6ietd he had 6ouncl 6 o 6a 6 ciFl a Ung in wa 11 Ume •••• EO 3 3b 3 PL 86-36 50 USC 3605 Vf LMAR C LANG •Ch e 6 80 4 • MJt Lang llpen t 23 yea Jtll ln the Ai1r Fo1r ce 16 06 th em w-Lt h the USAF Sec 11 l ty Se1r vice p11 - lo11 to 11 e tiJting in Augu 6 t -1965 He i6 a 1949 Chine6e language gJt dua te 06 the one-yeaJt AJt y Language Sc hoot c ouJt 4t and wa 4 in6 t 11 u me tal in e6 ta bti6hment 06 - the 6pec ia lized Chine4e Language T1r ining PJt ogJt am 6oJt USAF S ca11 ee11 in· the SlGINT Highlig u 06 hi6 community nctude 14 month6 in Ko ea n 1952 53 du Jting which time he pionee 1ted the ' 06 S G Nt in •uppoAL 06 LacUU t aiA opeAation• a6 066ic e11 -in-Cha11 ge 06 the Chine6 and NoJt th Ko11 ean 8Jtanch 06 the AFSS Eield PJtoc e64ing Cen te11 15 month4 a fl Ope a on6 066iceA aL USA-57 duA ng which Um€ he • uadAon ••Labt •hed the ope aUon which became USA-69 at_ _ Ia tou all the B oup 8 S ta 66 Repll e4en ta ve a Hq NSAPAC Camp Fµch nobe Jtpan and a tou a Chie6 NSAPAC RepJte en ta tive rOUJU I - Hi6 a64ignmen t6 at NSA have encompa46ed va11 ying ta4k6 in 83 including a 4tint all Vepu ty Chie6 Oeputy Chie6 BS and Chie6 BOS 6 om 1963 to late 1967 In the latte11 a 6ignment he wa deeply involved in the application 06 S1GINT in 6uppo11 t 06 tactical 6011 c e6 in Sou thea6 t A ia 3 'f6P SECRE'f UMHttA TOP 8 BORE'f UMflltA EO 3 3b 6 PL 86-36 50 USC 3605 JOHN 8 CALLAHAN Ch le 6 B05 John B Callahan'4 involvement wi h SlGlNT 4pan4 24 yeaJt4 and thJtee con Unent4 It 4taJtted in 1948 wlth hi4 militaJty a64ignment in the A 1tmy SecuJtity Agency a ·HeJtzo Ba4e GeJtmany and continued when he joined NSA in Ma1ch 1953 a4 a civilian tJta66ic analy4t and JtepoJtted on the S C viet MilitaJty pJtoblem Th 1tee yeaJt4 lateJt he wa4 bac in E ope a6 analy t and con4ultant with C1FCO the A 1tmy CentJtal zed P 1tog 1tam Hi4 JtetuJtn to NSA and the Soviet I IPJtoblem came in 1959 and 6oJt the next 6ix yeaJt4 MJt Callahan held va 1tioU4 SIGINT JtepoJtting and con4ume 1t Jtetation6 po6ition4 with PROV GJtoup A To highlight thi6 peJtiod he helped e4tabli4h and maintain the GJtoup A Watch CenteJt in Jte6pon4e to the Cuban CJti4i4 06 1962 SeptembeJt 1965 6ound him detailed to the VIA Intelligence SuppoJtt and Indica Uon4 CenteJt wheJte he 4pent a yea It inteJtpJteUng SIGINT mat teM 6oJt thi4 majoJt U4 e 1 M 1t Callahan'4 SIGINT attention 4hi6ted to the Fa It Ea4t in SeptembeJt 1966 with hi6 M6ignment M Chie6 Intelligence Sta66 GJtoup 066ice 06 Communi4t Sou thea4t A4ia A natuJtal development 6Jtom that job Wa4 a move to Vietnam wheJte he pent anotheJt yea It pJtoviding inteJtpJtetive 6uppoJtt 06 SIGINT pJtoduct at VoV Spec Rep MACV Sack once moJte at NSA he became Chie6 6iut 06 812 SEA Non-Communi6t Nation41 and then 06 811 KoJteal He M4umed the po4ition 06 Chie6 B GJtoup Intelligence Sta66 805 in JanuaJty 1972 4 T6P SECltET UMHlb EO 3 3b 3 PL 86-36 50 use 3605 'fOP SECftE'f UMBRA MAYBE IT'S RELATED TO THE PHASE OF THE MOON by Herb Guy 845 This story of the dissection of a callsign system proves the validity of that old saw Many a true word is spoken in jest It proves a lot of other things too--among them that it ill behooves the cryptanalyst to dismiss the word spoken in jest too quickly But you may ask what a cryptanalyst is doing dissecting a callsign system in the first place--isn't that a job for a traffic analyst Well in case some of us haven't yet learned the lesson that you can't really draw a line between the work of the cryptanalyst the traffic analyst and the linguist this story provides a bit more proof of that too The reader has probably guessed by now that the title of this piece was the true word spoken in jest But it wasn't reallv scoken entirelv in est because we knew that manv of s 'fOP SBORB UMBRA EO 3 3b 3 PL 86-36 50 USC 3605 'JQP 8ECKET UMRKA 6 'f6f Sft ltE'f UMBltA -· --···- - - • --- ' - - ·-- - -- -- -·- --EO 3 3b 3 iOP St Cftt 'f UMHltA 1 PL 86-36 5 usc 3605 7 TQP '3 Bm UMBlb I EO 3 3b 3 PL 86-36 50 'fOP SHCtffi'f UMBlb 8 T9P SECM'f UMBRA use 3605 EO 3 3b 3 PL 86 - 36 50 i'OP StiOM'f UMBRA 9 OP SECRE'f UMMtA - - -- - - - - - - - -- - - - -- use 36 05 EO 3 3b 3 PL 86-36 50 USC 3605 10 'EOP GC n' UMBRA TOP S Clffl'f UMBR A 11 SEGAR UMB ' ' EO 3 3b 3 PL 86-36 50 USC 3605 EO 3 3b 3 PL 86-36 50 USC 3605 12 TOP 8 BOR B'f UMBRA THE REALITY OF COMMUNICATIONS CHANGES by E E Orr B41 All analysts and managers of analytic efforts must constantly face both the possibility of a communications change on their targets and the consequences of such a change The term communications change frequently causes unnecessary apprehension--the change does not inevitably signal adverse consequences on target identification maintenance of continuity and production of SIGINT Many changes introduction of new callsigns frequencies etc on most targets are routine they occur regularly and are only slight hindrances to the proficient analyst On the other hand some communications changes are not routine and do have an adverse effect on SIGINT production They can result in reduction or even total loss of capability to identify and maintain continuity on target communications nets and the specific associated terminals The latter type of communications change is the subject of this article Changes which might affect exploitation capability will vary greatly for different targets depending on the extent of current exploitation and on the complexity of the newly introduced operational procedures However knowledge of the relationship between various communications features can greatly assist in prediction of future operational usage Some features which should be considered follow 1 SOI life expectancy Most signal officers are systematic in their Signal Operating Instructions SOI and are apt to practice cyclic introduction of new materials such as callsign systems Knowledge of their idiosyncrasies helps in predicting the extent and date of a change In any event operating materials which have been in use for extremely long periods are more likely to be replaced than those recently introduced 2 Cryptographic continuity Past experience shows that a change in such operational communications procedures as callsign or frequency usage is not usually accompanied by a change in the cryptographic procedures applied to either valid or non-valid text--probably because different organizations and personnel are involved Usually the cryptographer does not 13 0P SECRET UMHB1 -------------------------- ------- EO 3 3b 3 PL 86-36 50 USC 3605 transmit the message and the transmitting operator is not aware of the method of encryption Thus textual characteristic$ can often be exploited in lieu of external characteristics and vice versa 3 Sudden versus gradual change Many changes e g newly allocated frequencies can be implemented immediately upon receipt Other changes require live testing and extensivj operator training and orientation The following changes for • example would probably require an extended period for implementation a Introduction of a more sophisticated mode of• communications Equipment procurement is usually limited and testing and training are required before the new system becomes• operational • b use of a new Morse cut number system Operator training is obviously required prior to full implementation c Introduction of - --r----- --- - ---_ -- --_ --- ------_ Ls of a c ange requiring extensive opera or raining an examp e Some indicators of an impending communications change are 1 Temporary extension of the normal period of use of existing SOI materials 2 Limited testing of new procedures on existing links nets or on supplementary communications 3 Direct references in chatter to new procedures Such references could consist of anything from a casual implication to a statement of the effective date and type of new SOI materials 4 Trends toward standardization or diversification whichever is applicable 5 Use of or references to more sophisticated modes of communication 14 TOP SEORB'f UMBRA ----- Although the ability to predict impending communications changes is a distinct advantage recovery of continuity on target communications is greatly expedited by contingency planning which defines actions to be taken following introduction of new SOI materials Contingency planning in preparation for subsequent analytic recovery must be realistic and flexible Consideration should be given to the following factors 1 Timelf field station reporting of deviations from the norm As the mission of most collection sites is limited in scope this reporting permits higher echelon to make an early assessment of the overall extent of the communications change to advise all elements concerned and to issue necessary instructions 2 Tar et recognition identification Even though such things as ca sign and frequency usage have changed the best source of target recognition identification is the operator who has copied the target in the past and who will probably recognize it in the future Operator identifications should be considered valid unless disproved These identifications should be provided in a format usable for traffic identification to other field sites which are tasked with similar targets and which are likewise encountering difficulty in isolation and identification of mission targets Thus time will not be wasted in copying communications which are another site's mission 3 Establishing procedures for early continuous follow-up collection on potentially mission-associated communications Although these communications may not be identified beyond nationality establishing procedures for early collection will prove most advantageous 4 Determining possible methods of attack as a means of associatin homo eneous interce t and erformin follow-on ana sis In making t i s eterm1nat1on we must as at woul we do if the old tried and proven analytic techniques and aids were no longer available A definitive answer to this question will probably not be found but alternate approaches can be devised For example if callsigns cannot be exploited related intercept can often be associated on the basis of cryptographic features These features may therefore need examination very early after a communications change 15 TOP 8 € UMBRA ' rI 5 Once the possible methods of attack have been determined develoaing detailed erocedures for quick implementation These proce ures include issuing instructions to be followed in the event of a communications change outlining processing preferably in conjunction with a flow chart and devising the machine software which would be needed for machine processing Processing of data after an extensive communications change does not require completely new procedures although some alteration or expansion of existing standard procedures will probably be necessary Maximum retention of established procedures which are already well known to all operating elements will cause minimum confusion following a communications change and will aid in early recovery 6 Maintaining continuous documentation on all special processing or analytic actions taken and the type extent and data of actual changes in target SOI This documentation will aid in keeping all elements currently informed and in preparing for later SOI changes If this article succeeds in stimulating more realistic planning for future communications changes deterioration of SIGINT production after such changes will be minimal rANi f l I 1 i ___ • he gem cannot be polished without friction nor man perfected without trials 16 TOP SECltET UMffftA - - · • · - · - - - - - - - - - - - - - - - - - - - - - - - - A NEED FOR A CENTRALIZED TRANSCRIPTION OPERATION by Richard S Chun B44 It is well known that the introduction of new and more sophisticated voice communications facilities by B target countries is expected to produce a corresponding increase in the volume of voice intercept It is also well known that the shortage of transcribers both in the field and at NSA will become increasingly critical if we continue the present concept of voice operations B's problems are even more exacerbated by the fragmented and diversified voice transcription setup now in effect Three voice processing laboratories probably four after the transfer of F44l's mission and functions to NSA in June 1972 are managed operationally by the several B operating elements but the equipment accountability and maintenance is the responsibility of B44 This results in varied and parochial processing and reporting procedures training doctrines priorities and records and files maintenance systems Further experienced transcribers assigned to elements which require little transcription work have moved to more lucrative career fields thus producing the current feast-or-famine transcription resources situation in B Most of these problems could be solved by having B's voice transcription operation under a single management at a single location A centralized voice transcription operation which assembles in one unit the career-minded and professional linguists would help ease the acute shortage of transcribers linguists since the experienced linguists can be cross-trained to process any communications entity Bat present has no documented standards for consolidated voice tape accountability disposition records intercept requirements priorities RT handbooks training aids standardization of terms training doctrines or other data necessary for an effective and efficient total voice transcription operation These requirements can best be met under centralized management Ten or more steps are presently necessary to process a single multichannel tape from intercept to degaussing i e erasing -not including the numerous other steps performed by the analysts handling the same tape before it reaches the OPI A centralized 17 'JX I MJRIR UMfHtxA - - - - - - - - - --- ----- I I I l iI· f I Iri ff P SEORIB' UMBRA effort would limit these steps to intercept demuxing transcription translation and forwarding of processed material to the OPI analysts Other advantages that would accrue from a centralized transcription operation follow B OPERATIONS a Adjustments can be made to loss of transcribers changing field transcription capabilities and shifting requirements b A central control for voice-related technical services e g signal analysis data processing etc technical support to field operations coordinating effecting voice intercept equipment accountability maintenance and operational quality control and for voice-related research and development including special projects c Establishment of standardized voice transcription processing and reporting procedures formats a single operational training doctrine including SOT OJT and intern programs and a centralized voice-related language research effort ADMINISTRATIVE AND HUMAN FACTORS a Elimination of administrative redundancies under the single management and better long term programming and planning space personnel equipment b More opportunity to increase transcriber language capabilities by offering greater variety of assignments and improve transcriber morale with better career planning professionalization 18 'fOP SHCfttJ'f UMBRA TOP SECRBCf UMHRh EO 3 3b 3 PL 86-36 50 USC 3605 i •f •i THE OPEN DOOR _ -· • 4 • t • t ' i We aeek to be companion aton the way The lante n which we ca ky ia not ouka The api it which we ahake ia contagiou thought The knowledge which we gain an illuminating ' I ' • to 'tcf1 And all who aeek may pekceive and teakn -The Concept 06 Vkagon Seeda • • THE ROLE OF MATHEMATICS INC A by Dr Ralph w Jollensten Pl There are often cases in cryptanalysis where mathematics is needed and is perhaps w ii1 w 1-- 6llilii 1ii11 w _ iii 1 1 W 1 __ 1 1 1 jj11 w _ i w _ ___ example the familiar These examples should make it apparent to all that the degree to which mathematics can be used in cryptanalysis depends upon at least three factors 1 the nature of the C A problem 19 f9P SECBBT UMBRA 'fOP SECltE'f UMBKA EO 3 3b 3 PL 86-36 50 USC 3605 2 the mathematical background of the analyst and 3 imagination and cleverness with which the i a ilf ao aa - - ----- his backaround knowled e to the roblem Perhaps the most important factor is the imagination and cleverness with which the analyst can apply his mathematical knowledge to the problem Analysts often look at a C A problem and conclude that mathematics is not applicable to the case Mathematicians are often hired to fill C A intern billets and soon bemoan the lack of opportunity to apply their skills C A interns take probability and statistics courses programmed by the C A Panel and upon completion are asked To what extent does the course apply to your job In many cases the answer is Not at all I believe that in these situations the main reason the analyst cannot see an opportunity to apply mathematics to C A is a lack of imagination or desire rather than a lack of experience Non-mathematicians are frequently stymied by mathematical symbols and notations and hence shy away from its use while mathematicians who may be inexperienced in cryptanalysis often attempt to apply their skill 1001 of the time whether or not it is required It is often difficult for an analyst to compile a mathematical formulation applicable to any problem much less a cryptanalytic one Imagination intuition and patience are required in formulating mathematical problems One should not expect to become n efficient practicing mathematician overnight My advice to the young mathematician who bemoans the fact that he cannot apply his trade as much as he would like is this 1 Don't try to apply mathematics to every phase of the problem--an all-encompassing approach is often impractical Look for opportunities to apply different facets of the subject to different phases bits and pieces of the problem For example use counting techniques to compute work factors to see if a particular method will work in a practicable amount of time use 20 WP CMll UMBRA 't statistics to set thresholds use euclidian n-spaces as models in which to irnbed frequency counts use probability to compute the odds in favor of one hypothesis over another 2 Don't insist on using only your particular specialty--algebra analysis or whatever it might be Be willing to look for opportunities to apply other facets of mathematics l 3 Read the literature available in our libraries on the application of various branches of mathematics to C A Become acquainted with specific cases which demonstrate the wide and deep applications of mathematics such as PTAH eigenvector techniques Fourier analysis and applications of polynomials over a mod 2 field My advice to non-mathematicians is this 1 Don't shy away from mathematics because you don't understand it If you are thoroughly familiar with the cryptanalytic principles involved the problem itself will help you to understand why certain mathematical techniques work 2 Don't let symbols and notations throw youi use your cryptanalytic ability to break the plain code used in the mathematical world 3 Make an effort to improve your understanding of the subject Especially concentrate on understanding probability and statistics and attempt to associate mathematical models to cryptanalytic problems A student once asked me why he should study the effects of rolling a die since we didn't run into dice in cryptanalysis I said In your homework which would you rather consider Rolling a die with six sides or a die with 26 sides The probability of seeing an A or any other letter from B through Z in flat random cipher would be 1 26 and a die of 26 sides is a reasonable model But the student could not see beyond the surface Finally for all analysts keep an open mind about the use of mathematics in C A and remember opportunities to use new mathematical techniques in C A creep up when you least expect them 21 l'OP Bl30R 1 UMBRA · -- a -----W-1L2£ £1 ' 1 '' _' CRYPTO-SCRAMBLE Richard A tki111 m Unscramble each of 1he five numbered crypto-scrambl11 placing one letter in each space to form five words or names each of which fib 1he •finition to its right 1 2 M A N F R E O A S Q U I R E ___Q _____QQ __ _ RAHRI Order of superiority in e Mt of whnls driven by notch rin91 CHEY _ Q __Q ___ _ 3 4 Rotor development table 2 wds ALLPALER Pair of wira of equal length ____Q _______Q N O F I T I T O N A C A R Encipherment procns involving encipherrnent disassociation and furthtlr encipherment DE ALS P E NT Stationary 11ts of contacts at the end of • __Q __Q __ _____Q __Q maze Now erran91 th• circled letters to form 1he cryptoanswer suggested by 1he cartoon at the right Print CRYPTOANSWER here HIPPIES' VIEW OF THE POPE Answer on page 31 22 t'OP 8tlCR B'f UMBRA 'ftW SBCRiff UMBRA MACHINE-AIDED TRANSLATION by Norman Wild B03 Machine t4an 6lation ha 6 be en di 6appoinu ng to opu mi 6 tli but i tti 6ailu4e to me a 6u4e up to acceptable human 6 ta nda 4d 6 6 houtd not lead U 6 to cLi 6mi 6 6 the whole thing a 6 a total lo 6 6 16 the ma chine cannot tu 4n out good t an 6lati on 6 it can p ovide welcome help to human t4an 6latoM In thi 6 and in tJAJo 6u b 6equent a ti cle 6 M4 Wild di 6cu 6 6e 6 the p4oblem 6 inhe4e nt in t an 6lation the u 6e 06 machine 6 to aid t4an 6la ti on and the hi 6to y 06 6uc h u 6e in NSA Problems of Translation In considering translation two classes of problems are evident those which are fundamental and which would exist whether the translator is a man or a machine and those which while they are of lesser magnitude still create difficulty especially for any machine-aided translation The fundamental problems are two 1 Language texts are often ambiguous they are open to more than one translation into the target language By this I mean translations which differ in substance not merely in stylistic choices Many of these ambiguities are resolved by knowledge of the real world rather than by mechanical examination of immediate context The human translator can usually tell whether x iang means elephane' or photograph or whether c hin 6 e n means wages or sunken ship If he is translating English into Russian he should know which of three Russian words to use to translate Pole t riot in Gdansk Lineman injured in fall off pole 11 or To the pole with Peary It is hard to imagine a finite program that would put all the necessary background information into a machine - 2 Languages do not correspond one-to-one A word or phrase in the original language has some fit with a word or phrase in the target language but they do not match exactly The translator has to decide which word to choose 23 OP SBORE'f UMBRA ------------- ---- -- -- - - ---·'W a -·---- when none is exactly right and several are partially right what to omit because it cannot be carried over into the target language and what to supply even if the original language did not indicate it For example the original language may have omitted the subject and tense of the verb which are necessary for an English sentence The translator is not looking for the one right answer but for the closest answer and different answers may be used for different purposes Such choices are difficult to leave to a machine prograrn--but perhaps not impossible Some lesser problems create difficulties especially for machine-aided translation 1 The text must bet or ke - unched with a very hig egree of accuracy Te mac ine oes not correct errors as readily as the human being If the text is not in Roman or Cyrillic or if it has special characters a cumbersome arbitrary coding may be necessary In extreme cases a satisfactory input may take more time and require a rarer skill than the translation itself Of course it helps a lot if the text is already in machinable form for other reasons 2 The target language may have highly inconsistent usage Consider for example the English usage ln Ireland on Cyprus and at home for the same locative meaning Unidiomatic choices by the machine can add an element of confusion or at least of unfamiliarity which slows down comprehension Feeding all usage in would make for a very cwnbersome program 3 Findin the base form re uires a reat deal of anausis and programming or a large burdensome voca u ag ist If e program is not designed to isolate and identify t e base form--roughly speaking the form under which a word is listed in a dictionary--all possible variations have to be stored Consider two numbers three genders and six cases of Russian nouns Even then finding the inflection at the end of a Russian word and allowing for ambiguities and irregularities is easier than finding the base form in a language which modifies the base in other ways e g n lt lj lum lza I hurt myself from -um la to be hurt 24 T6P SECft B'f UMBlb ------------ - ---- ---- ·-··--------- CfOP SFJORM UMBIM l l 4 The unit of translation may not be neatly set off by white space It may be part of an unbroken stream of syllables or even of letters from which it must be extracted and there may be more than one way to divide the stream 5 The unit of translation ma as in e 1 e xxx an We e xxx q- J- a n 11 for the purpose of xxx 11 what did you xxx for 11 or why did you xxx The translator must hold the first element in memory until the second element is found The second element may come much later it may not come at all in which case the left-hand element has a different meaning or when it does come it may be coincidental and not belong with the first element People handle this situation better than machines Vt b11 a c hte xxx u m 6 The contextual clue ma be far separated from the ambiguous word rather than immediate y adjacent For example a man's name may be spelled in full at first mention in a Japanese text and thereafter throughout the text or even in other texts at a later date the name will be given in a drastically abbreviated form 7 People respond better than machines to nonce-words or nonce-usages These are words or usages which never existed before but have been coined for an immediate purpose For example any foreign word or proper name could conceivably occur in a Japanese text when appropriate in a distorted form The translator who sees a pa11 utoha ito for the first time in a context dealing with South Africa should have no trouble reading it as apartheid He should not even be bothered much by hanpa t6 u literally repulsion as backlash in a Japanese discussion of the American election campaign of 1964 A finite program could not predict all possibilities and enter them in advance New meanings for old words must be caught semantically from the context or at least it must be realized that the old meaning does not hold Machine translation presents a special problem when a closed group of correspondents who share a context use abbreviated or distorted language which is clear to them but baffling to outsiders Vocabularies whether general or technical are larger than people realize The suggestion of a micro-glossary to contain only those words which will occur in the text to be translated brings to mind 11 If I knew where I was going to die -I wouldn't go near the re • 11 25 'fOP StlCMl'f UMBlb - - - - - -- - - - - _• · - - 4 A more modest goal than machine translation is automatic look-up In this operation the machine program finds words or units of look-up in the text finds the target-language meaning on a dictionary tape and prints the meaning Automatic look-up saves a lot of time but it could be a dangerous tool for the translator In the next article we will look at some of the pros and cons for the use of automatic look-up •••• ODE TO A VIETNAMESE CRYPPIE Minnie M Kenny B03 Last night as I drifted to sleep word to my conscious did creep All night it stayed with me it just wouldn't leave The word my dear Dunc was eceive A To work in the morn I did fly On my worksheet a pattern I spied I'm excited delighted relieved The word my dear Dunc was eceive So little by little it's read The wheels spin around in my head Next comes oo Look Here's o om Ohl what glee l Why yes There's the word T i-Pa ty Now put up your pencil my dear The matrix's too long to put here Don't worry or fret The next one you'll get The language Cambodian I fear 26 'fflP SECltET UMBRA - ------· ____ __ EO 3 3b 3 PL 86-36 50 USC 3605 STUDY OF ZFK MESSAGE ACTIVITY CHICOM y Kenneth Miller B 3 r 27 'POP SBORB UMBK1• · • 110P SF30Mlf UMBlt A 28 'fOP SECH ffl' UMBlb EO 3 3b 3 PL 86-36 50 USC 3605 EO 3 3b 3 PL 86-36 50 use 3605 OP SHORB ' UMBRA · 29 TOP SHCR B'P UMBRh - ----------- --- · - ---- ----------- 'JQP SMJRFJ UMBRA 30 'f6P SEClffi'f UMBRA --- · - EO 3 3b 3 _ P_L _s6-- 3-6 -5 o_ us_c _36- o5_ EO 3 3b 3 PL 86-36 50 USC 3605 ' POP SHCRE'f UMBRA ANSWERS Answer to Crypto-Scramble 1 2 3 Friedman Square Parallel 4 5 Fractionation Endplate Hierarchy Cryptoanswer Latin Square Answer to puzzle sent to the Dragon Lady 1 2 3 -1---1--1-1---1-- - - - - - - -------- ' SISSIES TOMTOM TARTAR ----- - -- - TOP SBOR B UMBRA VIETNAMESE COMMUNIST TACTICAL COMINT OPERATIONS 1'CSUt by Tim Murphy B6 The Vietnamese communist COMINT effort in South Vietnam is quite extensive but very much decentralized Its purpose is simply to gather and disseminate tactical intelligence on a timely basis to Communist units in the field In short the enemy's COMINT elements function in much the same way as U S and Allied direct support units do Their COMINT units are usually organic to fronts divisions or equivalent organizations and all tasking processing and reporting appear to be done at that or a lower echelon The Communists have no NSA-type organization in Sou h Vietnam Vietnamese Communist COMINT wiits are generally comprised of a number of mobile intercept teams and an element which has responsibility for processing and reporting the collected information Often these intercept teams are attached to units on combat missions in order to provide direct support They are tasked by the Intelligence Section of the parent organization's Military Staff through the COMINT unit's headquarters Typical of Vietnamese Communist COMINT organizations is Front 4's COMINT Unit 508 This independent military intelligence unit which is directly subordinate to the Front Headquarters operates in Quang Nam Province of South Vietnam Our own exploitation of Communist communications has disclosed that COMINT Unit 508 conducts an extensive radio intercept processing and reporting effort against U S South Vietnamese and South Korean communications COMINT Unit 508 has been remarkably successful in exploiting Allied tactical communications Intelligence information frequently reported falls into several categories such as the disposition of Allied forces requests for support e g air artillery medevac Allied after-action reports Allied intelligence activities future plans and VIP activity Their reports are quite timely and have contained information transmitted by the Allies slightly more than an hour earlier The mode of operation is for the intercept team to forward the collected data to COMINT unit 508's processing element Team 1 There detailed intelligence summaries--in many 32 10P OR8'f UMBRA EO 3 3b 3 PL 86-36 50 USC 3605 f6P EffiCRti'f UMBR I respects similar to the daily sl1Illll a ie •i sued by U S collection sites in Vietnam--are preparrd•4nd forwarded A typical summal ' was passed on I _ Like other Vietnamese Communist intercept teams COMINT Unit 508's teams concentrate their efforts against targets which are most easily exploited i e Allied plaintext tactical voice communications In addition they intercept and exploit some ARVN encrypted communications The processing element provides technical support to the intercept teams via aperiodic technical messages These messages typically contain data on Allied callsigns frequencies and crypt system recoveries Intercept teams are at times requested to aid in recovering frequencies of various Allied units Vietnamese Communist COMINT units apparently do not have organic communications but rather share the communications facilities of their parent command Generally landline telephones or couriers are used for passing COMINT reports but both COMINT Unit 508 and the Long An Subregion's COMINT unit west of Saigon use radiotelephones extensively As a result details concerning the composition and modus operandi of their COMINT organizations are available 0 the 36 way-6 to 6 i 9ht the but i -6 to 6 te e Old Chine4e P ove b 33 l'OP B ORB'f UMBRA · -- 4 - _ _ _ 'fOP SBOH'P UMBRA 1 SEEDLINC oil 1 - _ 1J ----What with the reorganization of B6 and the relocation at FANX of Bl you are reminded to update your organizational telephone directories to reflect related changes Also about calls to FANX don't be discouraged if you are disconnected amid streams Just hang up and call again It seems that the modern convenience of efficient communications links has not reached that new-world outpost yet A bit ironic considering the business we' re in educational opportunities in fields related to their assignments available as well Typical assignments for fellowship study are identified for area specialists cryptanalysts electrical engineers computer scientists mathematicians cryptologists and historians Applications should be submitted through supervisory channels Questions on the program will be answered by Mr Walter P Sharp 8051 or 796-6334 SECRET -------- ----The establishment of the Cryptologic Education Fellowship Program was annoJ U ced in a memorandum from the Commandant of the National Cryptologic School dated 17 December 1971 The program is open to civilian and military employees with a broad background in cryptology or related technical fields Those selected will be assigned to the NCSch for a year to participate in the development of training programs and in teaching with EO 3 3b 3 PL 86-36 50 USC 3605 ----There has been qu te a bit of interest in Bll03's Project BABEL since c•rol Leve P26 touched briefly on its conce t and i clud d por ion n er s p e e c--e-o r e the Book break ers Forum in April Using the Stromberg Carl on 4060 Plotter B1203 has 34 TOP SECRlft' UMBRA ·- - - - - -- -· EO 3 3b 3 PL 86-36 50 USC 3605 • • -- - - --------- ----- ----- op SHORlff' UMBR A • in an effort to furnish the l I A discussion of the techniques used will appear as an article in the next issue of v agon Se d ----The National Cryptologic School now offers a course in the diagnosis of manual cryptosystems as part of the cryptanalysis curriculum Designed to be taken after com- · pletion of General Cryptanalysis CA-100 which stresses cryptography and exploitation of known cryptosystems Practical Diagnosis CA-260 teaches the diagnosis of unknown cryptosystems Diagnosis is presented as a 5-step iterative process gathering the a priori information separating the undifferentiated traffic into homogeneous sets analyzing the sets and making hypotheses testing and proving the hypotheses and writing a solution report A popular feature of the course is that the student acquires his skills by diagnosing a variety of operational foreign language problems from A B and G Groups Statistical and computer techniques are emphasized This course is scheduled to be given only once in FY73 from 18 Sep-17 Nov Successful completion of CA-100 and an entrance exam are required ----Pl6 has as part of its mission to maintain an awareness of the state 0£ language work throughout PROD It is thus required to know such thin s as the quality of language work performed by PROD linguists their need for working aids machine aids training and so forth Frequently Pl6 can provide or assist in providing such aids and at times to train linguists Several persons in Pl6 are deeply engaged in machine programs whose aim is to assist operational linguists in their daily work Staff linguists versed in Cambodian Vietnamese Burmese Thai and Chinese can be made available to operational elements to assist with special projects or with language problems during periods of unusually heavy activity For more information contact Mr 41Wrence Chief Pl6 ext 3957 ----Have you ever wondered how to apply machine techniques to traffic analysis This very subject is treated quite interestingly in a course offered by the National Cryptologic School TA-261 Computer Aid to Traffic Analysis introduces the student to various machine techniques and through practical exercises shows him how to use numerous machine outputs which aid in analysis Successful completion of TA-200 and MP-060 or equivalent courses are prerequisites 35 fflP SEClffl'f UMBRA 'f6P SHOH'f UMBRA ASK THE ORAGON LAVY Dear Dragon Lady I have just very belatedly got around to reading your Dragon Seeds - undated but the first and only issue I believe I especially liked the charming Foreword signed The Editors Printing format was attractive and limitation of articles to 2 or 3 pages tempted the reader into tackling even unfamiliar subjects I had a few other reactions I thought I might express for better or for worse 1 Cryptanal sis Through Functional Linguistics by D P Lenahan B22 Interesting even to a non-Vietnamese linguist Two questions come to mind What attention is paid to fre uency in this analysis It would seem that the tones would etray themselves by being of much higher frequency than anything else Are there variants for the tones If so this fact of course could partially conceal the disparity in frequencies but there are good ways to identify the variants · By the way is Janet King Wild's excellent account of Vietnamese Bookbreakin still being read It dates from the early l9S0 1 s but it 1s a thoughtful analysis by a brilliant articulate linguist and much in it will always be valid 6 2 Recovery of a Vietnamese Communist callsign System by Wayne Stoffel B03 His stress on the value of historical research is good and much needed The tendency to stop work on a superseded system before it is fully understood and to switch to the new system where there is less material available for analysis leads to much waste and frustration This can be seen in the case of code and cipher systems as well as in callsign analysis Historical continuity is vital to efficiency and in many cases even to the possibility of success Previous systems must be well documented and available to the analyst Furthermo e 36 'J0P 8ECIIE'r UMBRA 'fOP SfiCRfif UMBRA the latter should be required to familiarize himself with the past before being allowed to tackle a new problem 3 Chinese Voice Myers B4 1 Solution to a Dilemma by L St Clair The problem of limited language skill i•n field voice transcriber interpreters is a general one Even outside the Chinese voice area we accept the risk of erroneous field translations much too trustfully I believe this whole problem needs a good deal more attention with a view to some general solutions 4 The Creative Translator by Thom Glenn B61 An 5 excellent article well expressed Analyzation of Data by Richard Curtin Bll What's the matter with that quaint old-fashioned word analysis I choked on initialization too midway through the text Style and sentence structure leave much to be desired Some parts are downright unintelligible e g paragraph 5 6 I liked your publication as a whole Kay Swift G543 P S Having read-again belated-your Nr 2 I take back what I said about Dick Curtin - or at least I see some reason for the vague and awkward style Obviously I did not analyzate the datat Most of the articles are beyond my ken - or yen - but I found it useful to stretch the mind a little Mary D'Imperio's article was beautifully organized clear and well phrased as usual Mr Gilbert's comment on honesty in management evaluation was a fine strong ccy in the wilderness The old World War II evaluation check list diligence attention to pertinent detail · speed accuracy versatility initiative etc was at least a help in enabling the manager to point out strengths and weaknesses discreetly And it served periodically to remind 37 'fOP OR fi UMRBh T-QP l CJHs4' UMBRA both manager and employee of the desirable qualities on which the evaluation should be based It was discarded - perhaps because it was too much trouble or became so routinized that it was felt to be meaningless On a scale of Excellent - Strong - Good Fair - Unsatisfactor one woman was given a Good for Accuracy because she was in fact very bad She protested to the highest court available Nevertheless I think we are again in a rut and it wouldn't be a bad idea to dust off the old form - or a modified version of it - and see if we couldn't put a little more meaning into our performance appraisals Dear Dragon Lady Who spilled the ink on the code room floor On page 25 of issue 2 implies that your readers are familiar with Morse code Inconsequential Puzzle on page 22 implies that you think that your readers have time to figure out puzzles so may be you'd like to run a short Morse quiz seeing if your readers can make English words out of the following combinations of dots and dashes All we've done is leave out the spaces between letters 1 • 2 3 Conceivably longer strings of patterned Morse strings could be concocted but I think these three are interesting enough to hold your ditty-hoppers for awhile Answers on page 31 Harry G Rosenbluh Pl6 •••• Good judgment comu 6 ll om e x pe ue nce- - ua lty e xpeuenc e wh i ch wcu the 1tuutt 06 poolt judgment 38 1'0P BBOREi' UMBRh EO 3 3b 3 EO 3 3b 6 TOP StJCltt 'f UMBRA •· ' PL 86-36 50 USC 3605 CONTRIBUTORS RICHARDS CHUN Deputy Ch-le6 06 844 -l 6 o ijawa l l and tgan hi 6 c1typtotog lc e pe«ence in 1950 a 6 tne ii4 6t SIGINT e1te an t-lnguiht in the 6 leld He pe1t 001tme •a6 a l1r an 6lato1t b akb1teake1t and lnte 1r p1r ete1t and al 60 conducted -lnte1r 1r ogauop 06 NoA th Kolr ean « 6oneM o Wair He Aepo ted to NSA in 19oJ headed the _ 7 ' '1 601r a YeaJt and wa 6 1tea4li9ne jll lll llo - l_ a1 _ _ _ '_ _ _ _t M fl wh e1r e he wo Jtk e d on • Following hi4 conve 4 on o c v an n 1962 he 6i1t 6t 6e1tvei1 a 6 the Veputy Chie 6 06 827 now 811 and tatelt wo1tk ed on va1tiou 6 othe 1t PRC and NVN p1tobtem 6 he1te at NSA and at JSPC whe1te e initiated the 6iMtllcommunication6 -lnte1tcept 6Aom ACRP M1t commenda inciµd s - · A1tmy Chun' 6 Commendation Medal 4 n d _ Di 6 t i ngui 6hed M ltita1ty Se4vicee aoJt 61 e 01t£f He - 6 the de 6 lgneJt 06 6eve1tal analytic wo1t ing a-ld 6 including the Chun Wheel wh-lch -l 6 'being U6ed by A l1r and A lit Ve6en 6e analy 6 t 6 thJtoughout the woJttd and -i 6 a Ko1tean Chinehe Manda1ti n and Japa ne 6e lingu l 6 t • 1 h eq n ° ·t HERB GUY 8403 ha 6 6pen t mo 6t o6 hi 6 20 yeaJt4 06 cJtyptologic 6el v lce in P1 and 84 oil the i t_p1tedece4 601t ol gan-lzation4 He hM a B A 6Ji om the Unive Mit 06 Fl 01t-lda and a n M A 61t om the Un lveJi 6ity 06 M-lchigan both 1 t mathe matic4 In 1910-11 he attended the Naval Wa Jt College Although mo4t 06 hi 6 Agency expe«ence ha 4 been -ln c1t yptand t y4 l 6 he -l 6 al4o ceAti6ied a4 a Mathematician and a Special R 6ea Ji ch Analy 6 t • VR RALPH W J0LLENSTEN 1r eceived h 4 8 A in Mathematic 4 61t om Ha 1 Ung4 College NebJt a4ka h l4 M A in Mathema tic4 and Science 61t om the Unive' 6ity 06 N p-'1 a 6ka and h-l 6 PhD in Ma thema Uc4 6Aom the UniveMity 06 V-l lLg lnia lie ha-6 twentyone yea M expelr ience at NSA whe1t he i4 cu1t 1t ently Veputy Ch le 6 06 P12 OIL Jotten4ten hM al 60 4e1Lved a 6 the Execuuve 06 the CIA Ca1r ee 1t Panel and the head 06 the Sc-i ence4 Oepa 1ttme nt 06 the National CJt yptologic Sc hoot · EO 3 3 h 3 PL 86-36 50 USC 3605 39 i'OP SBCRfi'f UMBltA EO 3 3b 3 PL 86-36 50 USC 3605 KEN MILLER Cf A Te chn-i c i an in 84331 ha 4 be en w-l th Ns ·· 6 lnce 1965 leaving 6011 a tou11 w-i th the Ma Jl lne Co1tp6 1966-1 969 0 h l 6 6-l1t6t a6 6 i gnme nt he tackled 841'6 PRC call 6 lg p pbte m and then moved to 8432 the Re 6e a11 ch B llanch 06 the C11 yptotog c R akch Oivi ion H i• cukk • l lend Lua h tlen U to the PRC h lgh-g1t ade m i t i ta1ty _ _ _ • __ __ pit ob te m -i n 843 The CIA Ca 11 e e 11 Pane a 6 acce ple M1t · l e 1t '6 aJtt-i cle publi 6he d -i n th l6 -l66ue a 6 6ul6 lttme n t a6 a ba 6 i c 1t e qu-i 11 e me n t 6011 ce 1t t-i 6icailon a 6 a p1t 06e 66 lonat c ypta na ly6 t us TIM MURPHY 6603 had a wide va11 -le ty 06 inte tlige nc e xpe 1tie n a6 an A lit Fo11 ce 066 lce Jt be 601t e e nte 11 ing NSA a 6 a ivilian i June 1968 He completed the USAFSS Communication lnte tlig nce 066ice 11 Cou1t6e and the CY-100 p11 og1t am be 6011 e 6e 1tu-i ng with A SS in Be lllin with Hq 7th Ai1t Fo1tce in Saigon and i th Hq USAF at Fo11 t Me a de Flt om 1968 until h l6 11 e ce nt a 66ig me nt to B6 3 Tim wo11 ke d a 6 a civilian T1ta66 lc Ana ly6t and a Spec lat Re 6e a11 ch Analy 6 t on the VC Miti ta 11 y pJt oble m in 862 He uce ived h l 6 M A in ln te 11 na tiona l Re la tion6 ln 1970 61tom Ge o ge town Un lve Jt 6 l ty whe 11 e te n ye aM e a1t lie 11 he had be en a wa11 Jed hi6 B A in Engli6h • EUTH E EV ORR B41 e n tett e d the c tt yp totogic wo r ld -ln 1949 H l6 a 66ignme n-U 6ince that time have included S-0vie t Chine te Communi6t V le tname 6e and 1 011 e a n p1t obte m6 Jt an ng 61tom j j I _ He ha 6 6 e 11 ved n analyt c manage X a l lta66 poA Uonl A l clo6e a66ociaUon with PRC development continue 6 in hi4 cu Jt ent a64ignment a6 Acting Chie 6 841 which i6 conce ned with unidenti6ie d PRC communicat lon4 M11 011 i4 a 911 adua te 06 the Unive Mity 06 Ma11 yland and hold6 p1to6e 44ionat ce ti6icailon in T a66ic Analy4i4 and Special Re 4ea11 ch Analy4i4 NORMAL WILV 803 i6 one 06 the Age ncy'6 601t emo6t mulUlingui4u He ha 6 be en with NSA and p e de ce 6 60 agencie4 6ince Se pte mbe 1944 wo k lng mainly with Falt fa4te 11 n tanguage4 It i4 Jt e tiably e po ted that he ead4 STC like plain language M11 Wild'6 academic backg ound inctude 4 the 8 A 1939 and the M A in Chine 6e and Japane 4e 1941 61tom Columbia UniveMity He i4 the autho11 06 nume 1tou4 tingui6 t lc e6e ence and t a ln lng aid 6 within NSA and ha 6 tong be en c once 11 ned with the inteJt play 06 compute Mand language 40 'f0P ·---- - O UMRRA -- - - - - - ----- - - · Reinember - -