P L 86-36 UV0 l1 3 rnWVU 3 11 ---_ _ - LrWWV l I W l Wl lDl f W W I1 WlD J RUSSIAN SIGINT AND ELECTRONIC WARFARE J 1o oooo o 1 o o ANOTHER WORD ON AG-22 J S A VEXING AGENCY-WIDE PROBLEM Frederic O Mason Jr S MORE THOUGHTS ON QUESTIONABLE SIGINT 7 LANGUAGE SKILL FILE 'j i L 86-36 8 NSA-CROSTIC No 5 A- J S- - - - - - - - - 10 MECHANIZED LANGUAGE WORKING AIDS I i o o o o 12 MACHINE-PRODUCED AIDS FOR THE LINGUIST A J Salemme o 15 LETTERS TO THE EDITOR 20 WORK QUOTAS FOR SOVIET TRANSLATORS 21 THIS 9QCYM8 T C9 TM S C9B8 'lORB -fGP-5IEREf- MA ERIAl elMiMelIls BIRNSl II IilS HSlJLj' SSM III I the pt 60 88S 1il9 11661 aMlII 1 BII'I Ii J lJlIO NlMiI iI II I 9riIiMM o Declassified and Approved for Release by NSA on '10-'1 '1- 20'1 2 pursuant to E O '135 26 vl DR Case # 54778 'fOP SECRE'f Published Monthly by PI Techniques and Standards for the Personnel of Operations OCTOBER 1976 VOL III No 10 WILLIAM LUTWINIAK PUBLISHER BOARD OF ED ITORS Editor in Chief Arthur J Salemme 56425 Co11 e c t ion 1L - Cryptanalysis IC8025S 1 Language P L ---Jlc8955S Emery W Tetrault 52365 Machine Support jb3 S 1 Mathematics Reed Dawson 3957s Special Research Vera R Filby 7119s Traffic Analysis Frederic O Mason Jr 4l42s For individual subscriptions send name and organizational designator to CRYPTOLOG PI 'fOP SHeRET 86-36 DOCID 4009738 UNCLASSIFIED P L 86-36 Russian SIGINT Electronic Warfare Translated by anslator's note AS on The reluctance part of the Soviets to discuss their own activities in the areas of inteUige nce-gathering and e ectPOnic Jarfare 1 unU come as no surpr1-se to the reader II Recently however the Soviets have triedl ' to document at least superficial ly in l I I open sources some of their activities in these two sensitive areas The following is a translation of an article appearing in Voenno-istoriches kij zhurnal Journal of MiZita t y Histo t y No 3 1975 entitled From the Histo t y of Radioelectronic Wa t fare The authors General-Major of Communicatio s Tr ps Grankin and Colonel V ZmuvshJ d1-sCUSS the evolution of Soviet SIGINT and electronic wa t fa t e from a historical point of view Al'I' though the article is somewhat sketchy the authors do cite Ministry of Defense Archives as the sources of some of their information The most important point stressed by the authors is the total integration of SIGINT with I electronic WaPfare I I have taken the liberty of deletI ing two sections -- one section pertain-II ing to Lenin's views on radio as a prol paganda instrument and the other pertaining to the role played by non-Soviet i e Gernun U S etc SIGINT EW in World War II since only weU-known WeBte BOU e oited The ddetiO 1 of these two sections takes nothing away from the main thPust of the article II I I I I I I ' ---------1 I' I I I ' understands enemy's radio- Pre-World War I Developments The concepts of the radioelectronic struggle were expressed by the inventor of radio the I I JlI ----- - -- October 76 CRYPTOLOG Page 1 UNCLASSIFIED I DOCID 4009738 UNCLASSIFIED Russian scientist A S Popov In a memorandum dated 17 March 1903 to the Russian War Department concerning the establishment of radio communications between Varna and Odessa A S Popov noted that it would be preferable to run the radio line not directly from Varna to Odessa but from Varna to Sevastopol' to Odessa That would allow the radio line to be farther away from the Romanian shores thus making it almost impossible for anyone to monitor its operation or to interrupt its operation with the use of electromagnetic waves As we can see the idea of the possible conducting of intelligence and the creation of radio jamming was expressed for the first time by the inventor of radio himself and it was he who proposed measures to protect radio communications from intelligence-gathering and neutralization operations The first attempts at conducting radioelectronic suppression operations i e the use of electromagnetic energy as a kind of weapon against electronic systems occurred during the Russo-Japanese War It is a well-known historical fact that in 1905 in the Tsushima Strait area the Japanese light cruiser Izumi which was proceeding along a course parallel to that of a squadron of Russian ships informed its headquarters by radio concerning the number of ships in the squadron The commander of the Russian cruiser Ural having detected the Japanese radio transmissions decided to suppress them by using the radio set on board his ship This was reported to the Squadron Commander-in-Chief Admiral Rozhestvenskij who incidentally had categorically forbidden taking such action Nevertheless the commanders of the cruiser Izumrud and the torpedo boat Gromkij on their own initiative used their shipboard radio sets to suppress the radio communications of the Japanese naval vessels Thus measures to suppress an enemy's radio-communications system were carried out for the first time by Russian military specialists World War I During World War I an organized struggle against enemy radio communications was waged by almost all the combatants At the beginning of the war the forms that were most widely developed were the monitoring of the operations of radio means and the interception of radio transmissions Jamming for the purpose of interfering with the radio communications in the armies of the combatant countries occurred very rarely An important role in monitoring headquarters and troops was played by short-wave radio direction-finders that were created during World War I by Russian engineers N D Papaleksi and A L Mandel'shtam With the aid of the radio direction-finders it became possible to determine the directions to the emitting radio sets and to triangulate their location According to the number of operating radio sets it was possible to deduce the area where enemy headquarters and troop groupings were located as well as any changes or relocations Together with the intercepting of radio messages it was possible to determine the intentions of the troops the sectors in which the basic efforts would be concentrated etc In 1915 the armies of almost all the combatant states including Russia 1 had special radio-intelligence services which were equipped with radio-intercept equipment and radio direction-finders Omission Years Between the Wars 2 During the years up to World War II the development of methods for waging the struggle against radioelectronic means followed the course of intensifying the radio-intelligence effort radio deception and radioelectronic suppression Radio-intelligence derives information only from intercepting radio messages and fixing the location of active radio sets with the aid of radio direction-finders Therefore if broadcasts of false information are organized then it is possible to mislead the enemy and force him to take actions which will be beneficial for friendly forces Thus a second element in the radioelectronic struggle appeared -- radio deception In the Red Army great importance was attached to problems of deception with the aid of radioelectronic means and in a number of instances it resulted in major operational successes For example Marshal of the Soviet Union G K Zhukov discussed this matter in 1 For a Soviet view of Russian naval SIGINT in World War I see COMINT in the Russian Navy WWI CRYFTOWG May 1976 According to a former Soviet Communications Service chief Marshal I T Peresypkin from mid-1915 to early 1916 Russian Army radio-intelligence groups were established under the control of army radio-battalion commanders who were the senior radiotelegraphists Each group located at Army Headquarters was divided into two sections one section monitored enemy radio stations while the other engaged in radio intercept I T Peresypkin Voennaya radiosvyaz' Military Radio Communications Moscow Voenizdat 1962 p 56 T R H 2 Peresypkin points out that during the Russian civil war some of the reversals suffered by the Communists were due to poor COMSEC on their part In a similar vein during the Russo-Polish War of 1920 according to Peresypkin a Polish General Staff officer stated that Polish COMINT units gained enormous amounts of information because of poor radio discipline on the part of the Russians Peresypkin op cit pp 108 -11 0 T R H October 76 CRYPTOLOG Page 2 UNCLASSIFIED DOCID 4009738 UNCLASSIFIED his book Vospominaniya i pazmyshZeniya Recollections and Reminiscences In 1939 in the combat actions against the Japanese usurpers at Khalkhin-Gol the command element of the Soviet forces considered the decisive factor of success in the offensive to have been operational-tactical surprise For purposes of deception and maintaining the utmost secrecy concerning our measures the Military Council of the Army Group simultaneously with the plan for the impending operation developed a plan for operational-tactical deception We knew that the enemy was conducting radio-intelligence and was monitoring our telephone conversations So for purposes of misinformation we worked out an entire program of radio and telephone messages The conversations concerned only the construction of defensive positions and preparations for the autumn-winter campaign The radio deception was based chiefly on a code that was easy to break o Subsequent events and the entire course of our offensive operation showed that the special measures of misinformation and deception as well as other measures taken in preparation for the surprise operation played a very important role and the enemy was actually caught unawares 3 WopZd Wap II During World War II radioelectronic warfare was further developed and perfected It was waged by the armies of all the combatant countries At such time wide use was made on the fronts of all three components of the radioelectronic struggle radio-intelligence and radioelectronic intelligence radio deception and radioelectronic suppression 4 Radioelectronic suppression was the most important and most effective part of the struggle inasmuch as by simply creating special suppressing emanations of electromagnetic energy it was possible to disrupt the operation of the enemy's radioelectronic systems and means For the first time in history the Soviet Army used a combination of all three methods of the radioelectronic struggle in the Battle of Stalingrad In November-December 1942 the forces on the Southwest Don and Stalingrad Fronts intensified their use of radio intelligence After encircling the German fascists '6th Field Army the command element of the Stalingrad Front created an electronic jamming group in order to suppress the German radio communications This group had several powerful radio sets In order to derive intelligence from the radio means of the encircled troops to direct jamming operations against them and to determine the effectiveness of suppression the 394th Independent Radio Battalion was activated Simultaneously a special radio set was assigned to deceive the 6th Army Headquarters by operating with the callsigns used by the headquarters of Mannstein' s group of forces which was attempting to break through the encirclement That radio set received from the 6th Army Headquarters 86 very important radio messages The operation of the enemy radio sets was suppressed by tuning our radio sets to the frequencies used by the enemy radio sets and conducting meaningless transmissions when they started to transmit The established control over the degree of suppressing the radio communications as well as the testimony of captured generals and officers of the 6th Field Army attest to the exceptional effectiveness of the measures that were carried out A qualitative leap forward in the development of radioelectronic suppression as the basic method in the struggle against the enemy's radio communications occurred in the Soviet Army in 1943 On 17 December 1942 General Headquarters of the Supreme High Command decided to create special radio-jamming units Two specialpurpose radio battalions were immediately created the l3lst and the l32nd In 1943 and 1944 the 129th and l30th Radio Battalions were formed respectively These were the first units set up for radioelectronic suppression The creation of these units was necessitated by the waging of the constant radioelectronic struggle All the radio units took the most active part in suppressing the enemy's radio-communications systems and misinforming him The 131st Radio Battalion operated as part of the Northwestern Front and the 132nd as part of the 3G K Zhukov Vospominaniya i pazmyshZeniya Voronezh and Central Fronts 1943 subsequently Moscow 1969 pp 161-163 these uni ts were transferred to the 1st Ukrainian 4 There is strong but inconclusive evidence and 2nd Belorussian Fronts that in addition to his regular communications From 1943 through 1945 the radio battalions duties Marshal Peresypkin also controlled Soviet suppressed the radio communications of the military SIGINT activities throughout WWII For German fascist forces in the army-corps-division example see Heinz Hohne Codewopd Dipektop chain of command These radio units achieved a New York Coward McCann Geoghegan 1971 p very high degree of success in suppressing the 37 Hohne cites a publication of the Institute radio communications of the encircled groupings for the Study of the USSR Prominent Pepsons in The radio battalions set up a complete radio the USSR Munich April 1960 as his source barricade against the enemy's troops For exthat Peresypkin was in charge of military inample from 23 June through 31 July 1944 the telligence in addition to his position on the 13lst Radio Battalion participated in the BeloCommunications Directorate of the General Staff russian operation during the encirclement and from the summer of 1941 until 1945 T R H annihilation of groupings in the Vitebsk area October 76 CRYPTOLOG Page 3 UNCLASSIFIED DOClD 4009738 UNCLASSIFIED and to the southeast of Minsk This radio battalion worked round the clock disrupting 522 urgent and 1665 routine enemy radio messages S Subsequently this same radio battalion was given the task of completely suppressing the radio communications of the encircled Koenigsberg garrison and most important of preventing its command element from communicating with Hitler's headquarters The battalion was very successful in fulfilling this assignment During a 24-hour period in the assault on Koenigsberg the main radio transmitter of the encircled garrison attempted to switch to 43 different frequencies and everyone of the frequencies was suppressed The corps and division radio communications of the defending troops were also completely suppressed Then the main radio transmitter began transmitting in the clear the garrison commander's order concerning the capitulation of the troops 6 The commander of the Koenigsberg garrison forces GeneralColonel Lasch said at his interrogation As a result of the terrible artillery softeningup the landline communications in the fortress were put out of commission I had hoped to maintain radio communications with Courland the Zemland group and Central Germany However the effective jamming measures used by the Russians prevented the transmission of radio messages and my operations could not be coordinated wi th the headquarters of the Supreme High Command That was one of the reasons for my capitUlation radio communications For example in February 1944 during the encirclement of the German fascist forces in the Korsun'-Shevchenkovskij operation the headquarters of the 27th Army received intercepted radio messages which were being exchanged between the commanders of the encircled troops and the commanders of the troops outside the ring of encirclement The command element of the 27th Army decided one night to prepare all the powerful troop radio transmitters to suppress the radio communications of the encircled forces When the enemy began to attack we succeeded in completely suppressing his radio communications thus making it impossible for the encircled grouping to coordinate their actions in breaking through the ring of encirclement Thus during the course of the Great Patriotic War the struggle against the enemy's electronic systems by means of intelligence jamming with special emanations and radio deception found its further development in the Soviet Army In disrupting the work of radio communications wide use was made not only of special units created for this purpose for the first time but also the troop radio means Omission Thus electronic warfare was further developed and improved during World War II The scope of application of the means of radioelectronic suppression increased special units subdivisions began to be created and the tactics for their application began to be developed Postwar Period The l32nd Radio Battalion also operated just as effectively In March 1945 the troops of the 1st Ukrainian Front waged combat actions to annihilate the encircled garrisons in Glogau In the postwar period imperialism has unand Breslau The l32nd Radio Battalion having leashed many wars and armed conflicts Expedivided its forces and equipment into two rience shows that in each of them means of groups successfully suppressed the radio comelectronic warfare were used They were used munications of those garrisons During a 15for resolving not only strategic tasks but day period from 5 through 20 March 1945 the also operational-tactical ones 132nd Radio Battalion disrupted 358 radio transSimultaneously the armed forces of all the missions in the Glogau area 735 in the Breslau technologically developed countries began to area and prevented 2801 other attempts to esdevelop a fourth element in electronic warfare tablish contact 7 protecting the electronic means of friendly Later on the battalion's main forces and troops from the intelligence-gathering and equipment were directed at suppressing the jamming operations conducted against them by radio communications between the headquarters the enemy of the encircled 5th Army Corps and the 9th At the present time electronic warfare is Field Army headquarters a undergoing a new phase in its stormy developIn addition to the special radio units in ment As events have shown not a single the course of the Great Patriotic War the combat engagement not a single operation by standard equipment issued to troops was often any branch of the armed forces can begin or used for the purpose of suppressing the enemy's can be carried out without the broad use of the forces and means of electronic warfare 5Archives of the USSR Ministry of Defense collection 131 list 36683 folder 1 sheet 9 6lOid folder 1 sheet 17 'Archives of the USSR Ministry of Defense collection 132 list 328334 folder I sheet 31 BlOid sheet 32 1 lhas prepared an Aliilo talei1 Biblio I raphy of op ln-source materials in Russian which deal with SIGINT and EW Ia obtain a copy of the bibliography write or call PI CRYPTOLOG Editor 5642s October 76 CRYPTOLOG Page 4 UNCLASSIFIED UNCLASSIFIED P L 86-36 4009738 CONPIBENtFIAb UNCLASSIFIED f Fifteen years ago an intercept operator copied a radio signal onto six-ply paper He copied while the case was active and usually went back and put in comments when the case went down or paused The comments appeared not at the place where the pause occurred but at the place where they were relevant If the operator discovered 5 minutes into the copy that he was hearing a letter wrong he could go back and correct all the occurrences of the letter -- or simply tell the traffic analyst who was colocated with the intercept operator The traffic analyst at the intercept site could analyze the traffic in detail -- at a leisurely pace -- and could inform the intercept operator immediately of his conclusions He could even consult with the intercept operator to help determine what was going on in the traffic The traffic analyst at the intercept site was an important factor in intercept quality His close contact with the intercept operator speeded up the training of jeep intercept operators and also contributed to trained operators' morale and motivation by continually keeping them informed of what was happening The better intercept operators always had a great interest in analytic details of their targets Gold-flow problems caused NSA to look for ways to reduce overseas manning At first the technology was not available to move the intercept operator back to the States but it must have seemed a simple matter to move traffic analysts back Why they could analyze traffic at NSA just as easily as at the intercept site couldn't they AG-22 would give the traffic analyst his traffic on a real-time basis at NSA Budget problems caused NSA not just to move those traffic analysts back but to eliminate them altogether There was a set of traffic analysts already at NSA and why did we need duplicative traffic analysis The problem is this those two sets of traffic analysts were not doing duplicate work The analyst at NSA was performing long-range analysis buildin on the short-term work done in the field The analyst in the field was providing a needed service in short-term analysis and in intercept quality control and in intercept operator training and in intercept operator motivation Can AG-22 or any other computer system do that I think not i'J jf ' 1 r the past several months a problem has both interfered with and frustrated the best efforts of our analysts Happily it has now been'solved but it required a new concept in collection and a bit of special reporting not included in the routine TECSUM The problem was complicated by a lack of funds with which to increase the volume of coordinated data collected and by unverified reports of items collected elsewhere Each item collected provided an example of a term which fell clearly into either an A or a B configuration but there were also several occurrences of a C and one example of a 0' configuration Also collateral reports included references of configurations which could not be any of these arbitrarily designated X until examples are in hand Normal collection was on a very small scale at first with many items collected but not reported It became clear even before systematic dual collection was attempted that the system relied on rotors of different sizes all stepping at the same time but the resultant sequence could not be reconstructed from unrelated intercepts E8IlF I BEIiT IA October 76 CRYPTOLOG Page 5 CONPIBENtFIAb Continued next page DOCID 4009738 UNCLASSIFIED Recently it became possible to coordinate two collectors so that sequential items from the same series became available With no more than ten such paired items together with previously collected data it became possible to reconstruct the rotors and to establish the entire A and B' sequences Further it became clear that the C and D sequences were based on permutations of the A B rotors although it has not been possible with present collectors to accuStraight mulate enough data to reconstruct the C D sequence The final problem -- that of understanding the conditions under which the A B or the C D sequences are used -- was solved when data not previously required was added to the TECSUM The solution required that two cups be collected sequentially using one or two operators 2 pairs from the same vending machine and that the J J Poker Fun cups For Consumer Amusement Only -K K J J 8 Use for Gambling or Awarding Prizes May Be Kt Unlawful be marked in the order received And the final problem required that a distinction 1 pair be made between hot cups the A B set and cold J Qt J cups the C D set A cup from the A set is J 10 distinguished by having cards red or black in 8 BRRBR order and is always followed by a cup from the B set which has cards in BBRBR order K Q o and is always followed by a cup from the A set Both A and B end in suits o o which is a reversal of the C and D finai cards o o o The following pairs have been artificially We now know the full A B sequence and have constructed from the answer since the analysts devised a handy-dandy chart which will let us in an unfortunate excess of parochialism atpredict from anyone cup what poker hand will tempted to frustrate competition for the solution appear on the next cup by destroying their data However the answer As for rotors each card position in each set has been checked repeatedly at the point of inhas three four five or seven cards which re- tercept and these data are thereby validated peat in order Rotors would give this effect even if they cannot be retrieved from the data base we hang on our wall However our collectors don't normally colFollowed by lect C D cups and reports of flushes not pos- First member of pair sible with either A B or C D sets remain unconfirmed Solutions to these remaining and vex- 10 2 8 K 10 7 8 9 A J ing problems must wait for other analytic 8 K 24 Qt 10 gt Q Kt groups to solve A 8 10 94 J J 8 10 At Meanwhi e here are sam le data yo an col Kv J4 Q 9 Q At lect mo e 1f you w sh wh1ch are suff1c1ent to K K K Q 10 10 a solut1on The f1rst cup we collected arb1trarily designated as at position ADOI in the Q 8 10 J K Jt sequence was the hand 10 8 lOr 2 Kt A 104 Kt 8 K K 10f We also collected samples of various winning 9 10 ' 2 Af hands as follows Q 2 Kt J 8 J J 10t 4 of a kind K 9 2 Q Kt A 7 A g4nO 8 2 2 9 9 8 2 K 7 7 2 7 J 74 9 10 7 7 A 9 8 K g K Kit Kf 9 9 K K g 10 ' 8 10 lOt 10 Full house K 9 I o K 9 A A'I Jt A 9 7 A J Q lot t At 2 Q J 94 2 Q 8t 7 8 J K At Q 2 8 A Qf The problem for you -- and the author will buy one cup of coffee for the first correct answer -- is to provide the sequence position of the only cup having four aces K Solution next month October 76 CRYPTOLOG Page 6 UNCLASSIFIED DOCID 4009738 P L SECRE'f SPOI E '6E TN'II' TI '11 ''tIlEITI'oo'lE'' II T I October 76 o CRYPTOLOG o pa e I ' SECRE'f SPOKE EO 1 4 c P L 86-36 86-36 DOCID 4009738 CONFIDENTIAL L 86-36 LANGUAGE SKILL FILE The 6oUowing aJL ti c1 e appeaJted oJUg-i nall y n Field Information Letter June 1976 hM bee n - Ughfty Jte vMed and upda te d by e I alLthoM OdUCing an accurate inventory of the Agency's linguistic resources has become a major management concern in recent years In this context the term resources encompasses assets versatility and quality While it has been possible to assess language requirements with a fair degree of accuracy by using the billet requirements expressed in the Table of Distribution TO attempts to assess capability have been frustrated by the lack of verifiable data concerning incumbents of those billets Frequently supervisors I evaluations or even selfevaluations have been relied upon in determining skill level Taking inventories has also been complicated by the fact that some people with a language proficiency are no longer working in the language career field and some linguists have a proficiency in a number of languages even though they may be using only one in their current position After analyzing the problem it was concluded that the managerial tools now available -- including the TO certification statistics and language proficiency test LPT r sults -- could not be relied upon to give an accurate inventory We also concluded that an accurate picture of the linguistic capability of the work force would be gained if a language skill level was associated with an individual rather than billet or Career Occupational Specialty Code COSC As a resul t a Personal Skill Code PSC was established and is now being used to identify the linguist skills of all Agency personnel regardless of the job title they may currently hold The skill code has four basic elements that indicate 1 skill level 2 language digraph 3 method by whicQ the level was verified and 4 year the level was verified Additional elements can be added to these to indicate a voice-analysis cauability r To date codes have been assigned to all civilians whose background reflects lawguage training everyone coded at the zer6' m R N c level has been encouraged to be testep w 86-36 those langu g U ingthe skill codes and the Personnel Information Distribution System PIDS it is possible to retrieve language data on-line by language digraph skill level and organization It is anticipated that this information could be particularly useful durin isi i October 76 CRYPTOLOG Page 8 CONFlDSNTIAl DOCID 4009738 CONPIBEN'f'IAL L - _ _ _ _ _-- llthepeople coded at the zero and one levels might be of greater interest for retraining or for receiving additional training The system printout in response to these requests would give each person's name grade COSC and the last two positions of his job number which indicate the language he she is currently using if any and his her skill code The system also allows managers to make use of the versatility of their work force On a day-to-day basis a manager may only be interested in the fact that an employee has a proficiency in the language of the primary target he is working on however secondary language skills could be of significance in areas where there is an occasional requirement for someone with training in other languages or where there have been shifts in the emphasis placed on individual targets Similarly languages for which there is no operational requirement or a very small requirement are of interest largely because of the information provided on an employee's capa- bili ty to learn a language in a pa tlidd4r c family of languages P L 86- 3 6 The Personal Skill Code is still in a developmental stage and will be modified in future months to identify individuals who have not yet achieved certification but have passed one or more parts of the PQE and to identify individuals who have the capability to work with voice as well as written language materials Guidance from the Language Advisory Committee will be sought in determining when a language proficiency might be considered dOT 1lant and when it should be thought of as beyond revitalization A summary of the language skills currently represented in the field for individuals coded to date is given in the following table If you are interested in obtaining a copy of the Guide to the Language Skill File which gives examples of the output that can be obtained by using each of the programmed segments in the PIDS or if you have questions on the use of the file contact M33 8267 or 8276 SUMMARY OF LANGUAGE SKILLS REPRESENTEV IN F I Field UI'IU6 October 76 CRYPTOLOG Page 9 CONFIBEN'fIAL EO 1 4 c P L 86-36 DOCID 4009738 NSA-CROSTIC No 5 By A J S DEFINITIONS UNCLASSIFIED The quotation on the next page lJa8 taken froom a published 1JJorok of an NSA-ero The firost Letteros of the WORDS speLl- out the authoro's name and the titLe of the 1JJorok WORDS October 76 CRYPTOLOG Page 10 UNCLASSIfiED L 86-36 - DOCID 4009738 UNCLASSIFIED I October 76 CRYPTOLOG Page 11 UNCLASSIFIED P L 86-36 CONPIBEN'fIAL D9GJQ6-3 0097 38 LANGUAGE t' tneIl l L '1lO e to a d 1 et$1l ' -n l''''e ' g te Aat f r----------------------- 8'his brief paper' is intended to present some general comments covering my experience with the CAMINO Natural-Language Machine Files and other computerized natural-language working aids It has seemed to me that these remarks might be of some value or interest to others concerned with mechanized language aids whether as users or as designers In considering the design of a naturallanguage dictionary as in any data management system however sophisticated or rudimentary there are three obvious kinds of things th t need to be done a building the file from external source s b maintaining the file by additions dele- tions and corrections and a querying displaying or extracti g data I I from the file for day-to-day oierational use The attention of most computer specialists has centered on the tasks under heading a the ways of looking up rearranging manipulating and displaying data from all existing file This is admittedly the most interesting area conceptually and provides the most stope for inventiveness in software design and programming techniques Unfortunately however at 1dast in my ei perience the real problems invplved in the design of natural-language worki g aids reyo1ve around areas a and b and especially ar a a the initial building of the gtossary le The querying or displaying of data from a completed file has rarely presented anY problems in the CAMINO files or any other mechanized workin2 aid' I havo work d with I October 76 CRYPTOLOG Page 12 EO 1 4 c P L 86-36 CONFIBI3N'fIAI Iblli BhE oIl eeMW'f elhltrHlEh5 ellh'l DOCID 4009738 CONPIBEN IAL In maintaining the file once it has been established and has completed its initial rapid growth the file sponsor must see to it hat the file continues to be responsive to all his users and their needs in spite of frequent reorganizations of Agency personnel changes in missions and new developments in the external world of international events He must make a continuing effort to collect contributions of new terms and solicit constant feedback from all file users while searching for new sources of data that should be added to the file Now let's shift our attention to the question of how the tasks outlined above can best be accomplished at present and in the near future and why To most computer-oriented managers it seems bvious to the point of triviality that each sponsor who needs a mechanized language aid should do all the work himself preferably at a terminal unspecified If he cannot do all the work himself for whatever reason he should get somebody else also unspecified in his own organization to do it for him The hypothetical manager we are quoting sees no difficulty in any of this and considers the problem solved by his advice In certain specific contexts where on-line terminals have already for a variety of historical or practical reasons become the primary or only way of accessing and manipulating all data for a problem it makes excellent sense to put language aids on line too In these cases I the needs of one spec1f1c set of users are be1ngamply met by a complete system design that includes all the working aids they require There arestill however many scattered users who are and may remain for some time outside of these advanced projects For them a generalized or standardized conventional file-maintenance procedure involving periodic updating and reiussue of printouts has been an effective and economical way of 17ettinll the iob done 4 6 - hl c P 1 86-36 Faced with all these immediate pressures the senior linguis has little or no time energy or peace of mind left over to work on the dictionary at all let alone do the extra initial work required to mechanize it He certainly is not usually able to spend hours of his time sitting at a terminal which he is not routinely using for any other purpose keying in his own data or making line-by-line changes Finding someone else to do these things within his own organization is also apt to be very difficult in practice Everyone else who is qualified is just as busy and just as unable to take time to sit at a console for hours on end concentrating on the dictionary Helpers who are not qualified almost invariably cause far more problems than they solve Our hypothetical manager whom we have been quoting as a Devil's advocate has an immediate reply by which he again seeks to define away the awkward resource-allocation problem If the sponsor organization doesn't need the dictionary badly enough to assign the necessary resources to it obviously it isn't really needed at all so we can forget it Again I cannot accept this representation of the situation however useful it may be in simplifying matters for the manager 1 i L ------------lI October 76 CRYPTOLOG Page 1'3CONPIDEN IAL _ _ EO 1 4 c P L 86-36 WANQb VIA EUR81HliT CliANNEbS 8ltbY DOCID 4009738 CONPIBEN IAL 1 4 c L 86-36 ----------------- October 76 CRYPTOLOG Page 14 CONPIDHN'FIAI I NBLE VIA e6MIN'f EIIAI 'ELS 6riLY DOCID 4009738 'fOil SECRE'f Uf'fIBRA l l I i I J 1 1 October 76 CRYPTOLOG Page 15 CfOP SECRET UMBRA EO 1 4 c - P L 86-36 DOCID 4009738 qJOP SECRIH UMBRA October 76 CRYPTOLOG Page 16 tOP SECREt UMBR i iii EO 1 4 c P L 86-36 -- DOCID 4009738 'fO SECftE'f UI'fIBItA I i October 76 CRYPTOLOG Page 1 7 'fOP SBCRB'f UMBRA EO 1 4 c P L 86-36 DOCID 4009738 TOP S CR T UMBRA October 76 CRYPTOLOG Page 18 '-FOP SBCRE'-F UMBRA EO 1 4 c P L 86-36 DOCID 4009738 TOP SECRET UMBRlt f'l'8P SEERE'F lIllfblt pro srammed to store and process the informl tion into worn use and Inteaning The computers can retrieve specific material and provide printouts in both Greek and English which scholars can use ftJr compiling lexicons and key-word-in-coft text concordances Dr Brunner says the computer reql Jres about two minutes to find all the uses of a given word in the harf-million words of Plato's writings The old method would take I scholar several years to do the same research Dr Brunner says So far the preject begun in 1972 has entered 18 5 million Greek words on and the data bank is being u by scholars in such fields as literature philolOllY lin Research in the classics that once n ' al QUiCk Y a nd el tiden t1y frees 0e 1 involve SChOl dcating and r e ad llUistics history and philosophy took months or even years can now scholar of the task and allows him ing texts analyztng word use reo til devote more time to analytical ct lrding each word with its sen nce be accomplished in minutes with work acc ing to Prot 'I'heodo e context on a separate index card the help of a computerized data Brunner director of the Theswl' ls and filing the cards alphabetically New York Times bank Of the classical Greek language Linguae Graecae project at the Uni- The process 1$ usually long and tedious accordlng to Dr Brunner Eventually the system will store versity of Calill mia at Irv ine 5 April 1976 With the IrVine system Greek 90 million words to form the first The traditiOnal method of acorded thesaurus of the language cumulating information about a spe- texts are -punc ed and The ability of tbe c ompiJter sys- cific author's writing or all the on 'mllignetlc tapes tn a special code UNCLASSIFIED tern to provide basic resellrch mate- writings of a given period would The tapes are SC lnned by oomputers Classic Use of Computers Resear ch in Greek r ec October 76 CRYPTOLOG Page 19 TOP SECRET UMBRAEo 1 4 - - - - _ -_ __ - __ __o P L _ - c 86-36 _- C - -' ----- - -' - - ' - - - ' - - - _ _ _ __ DOCIO 4009738 '_- -_''_ _ _ SECHEl' UNCLASSIFIED To the Editor CRYPTOLOG I I article Integrated Analysts ror Asia A Cohesive Approach CRyPTOLOG August 1976 virtually admi s to sex discrimination in the placement of integrated analysts Wh le I am perfectly willing to admit to the desirability of empathy toward the Asian attitude I am also only too well aware of the large n ber of women whose lack of promotton has been explained by but the male candidate has had an overseas tour Does B Group need reminding that Agency policy forbids them to deny women desirable overseas positions on the basis of sex I think I lowes CRYPTOLOG readers an explanation ' brand To the Editor CRYPTOLOG Having spent a year as the Operations Officer of the 6924th Security Squadron Danang RVN I was most interested in the excellent article on IRONHORSE which appeared in the October 1975 is-iue IRONHO SE r ll t icaluu SIGINT System L Ju uThere was however one small but significant error The author's statement that the Marine bomb dump explosion destroyed the operations facility in April 1969 is correct However the system was back on the air on 21 July 1969 not 1970 as the article states I realize that three months to build an operations building clean up the entire IRONHORSE System install the intercept positions and IRONHORSE seem impossible Because of the outstanding cooperation received from the Commander 7th Air Force Director NSA Commander 366th TFW Danang RVN and the 6922nd Security Group and most importantly the cooperation and long hours of the REDHORSE construction units the USAFSS E and I team and the Operations Supply Com munications and Maintenance personnel of the 6924th Security Squadron such a feat not only became possible but was accomplished with smoothness and completeness in'less than 60 days The decision to rebuild the Operations facility at Danang was not received until late May 1969 Such an accomplishment further reflects what was alluded to in other areas of the same October 1975 issue as a certain sense of pride and professionalism in doing a job well no matter what the circumstances f------------J SE6RE'f 66El P S I'm surprised you let him print that I ' Edi tor' 8 rep ly uuuuuu u uuuuuuuuuu P L 86- 3 6 ThePublisher ditorial Board very r 1 fely advise an author to modify his or her statements on the basis of inappropriateness We eel that inasmuch as the views expressed in ea h article are the author's own and do not hav e to be coordinated or cleared with any6ne else at any level unless the author so esires every author has the right to express his or her technical opinion however controversial or even however half-baked it might seem to be lt is not true as people sometimes say around the CRYPTOLOG office that CRYPTOLOG will print anything CRYBTOLOG will print anything that has operationaL relevance If any reader disagrees with the author's views that reader is welcome to contribute a counterview anonymously if desired and that counterview will be printed so long as it contains a valid statement rather than just ad-hominem should we also include ad- feminam invective So in ffect I'm surprised that you're surpr sed that CRYPTOLOG let him print it Ed Speaking of his or her o To the Editor CRYPTOLOG I have admired the style of CRYPTOLOG since the first issue The quality of the English therein has been a delight to its readers and an example to all of us who write But I must protest the use of two particular pronouns in one recent article Is the use of he she and his her to refer tou a ' perator really necessary It seems to me that it condescends It is a repeated reminder of something we accept -- the EO October 76 CRYPTOLOG PaEe 20 1 4 c P L 86-36 SISCHE'f DOCID 4009738 UNCLASSIFIED operator might be a man or a woman Those who do not accept this will certainly not be influenced by seeing these words in print In fact these readers will probably be far more annoyed than I am If such usage is required by agency or governmental guidelines then I am alarmed If not then let us revert to the use of masculine pronouns to refer to a person of unspecified gender G43 Editor's reply As far as I know there are no such agency or governmental guidelines As far as CRYPTOLOG editorial policy is concerned we do not add or her unless the author himself or herself includes those words in his need I add or he I article Personally I feel that excessive use of his or her draw'S the reader's attention away from what is being said to how it is being said I would hope that most CRYPTOLOG readers do read the magazine for what its authors can communicate to them Unfortunately the English language itself is flawed People can suggest all-purpose pronouns that refer to either males or females but that doesn't mean that everyone who speaks English will accept them Even if one uses the graphic convention s he how would une pronounce it Moreover the suggested conventions aren't really that all-inclusive anyway A letter to the editor of The Wall Street Journal last year suggested that the all-inclusive pronoun should include not only he and she but also the third-person inanimate pronoun it should mean he or she or it The writer suggested the word h'orsh'it I think he's on the right track WORK IJUoTAS P L FOR SOVIET TRANSLATORS I fB5 Author's note The following information is derived from an article in The ATA Chronicle which in turn was extracted from Scientific and Technical Translations in Soviet Industry by I P Smirnov The All-Union Translation Center for Scientific-Technical Literature and Documentation in Moscow employs approximately 5000 full-time translators and several times that many part-timers These translators as all production workers in the Soviet Union have work quotas to fill In filling these quotas the translator must read the manuscript before it is retyped correct the retyped version proofread the corrections and closely scan the corrections made by the scientific editor No allowance is made for difficulty of the text And what are these quotas Languages are Ed divided into three classes with a different UNdASSIFIED quota in number of words per month to be translated for each class The word count is made from the target language usually Russian and not from the source language thus penalizing the translators -- e g no definite and indefinite articles in the count Occasionally the only comment the Editor Quota words hears about what he thought to be a particuper month Source language larly informative issue ofCRYPTOLOG is I 'There's a typo on page 13 11 Therefore he English German French 24 000-32 000 was pleased to see in a local Maine-coast Spanish Italian or a newspaper recently the nice way in which a Slavic language fellow editor handles the situation Hungarian Finnish Dutch 16 000-24 000 Portuguese Albanian Greek Romanian or a Scandinavian language 14 000-20 000 An Asian language If the translator dictates -- rather than writes or types -- the translation the quota is increased by 8000 words Any volunteers for work norms UNCLASSIFIED FIND ANY TYPOS October 76 CRYPTOLOG Page 21 PI-Oct 76-53-24865 UNCLASSIFIED 86-36 DOCID _ - - I urr 4009738 - - _- - 'K I fIllS BOCUMENf CONfAINS COBETt lORB PttA'fERIAL This document is from the holdings of The National Security Archive Suite 701 Gelman Library The George Washington University 2130 H Street NW Washington D C 20037 Phone 202 994-7000 Fax 202 994-7005 nsarchiv@gwu edu