lIJaJUDUJUJaJl1 17I WU l lBl llEDU aJ l UJlB l I W l Wl jJ l f W jJW I1 jJ Brn llU J b J1 APR-JUN 1980 86-36 5 DATA FLOW - CHALLENGE OF THE 19805 U oooooo Cecil J Philli 1 5 A TRAFFIC ANALYST LOOKS AT COMPUTERS U P16 LANGUAGE AND CRYPTOLOGIC LIBRARY U ooo oo AND IN A MORE MODERN VEIN ooo U ooooooo p- OH K UJ _ o o -- - - ----I NSA-CROSTIC NO 31 U ooooooooo oooooooooooo D H Woo ooooo ooo oo oo 10 LIME-A OHIO LEEM-A PERU U o oooooooooo o A J Salemme 12 AIT U o ooo ooooooooooooooooooooooooo ooo 13 SOVIET C3 U ooo ooo oo ooo ooo 16 WHAT TO DO ABOUT FANX U o o o ooo o oo o o o o oooo o oo 19 HELP WANTED U o o ooooo ooooooooooo ooo ooo Tom Engle oooooo oo oo oooo 20 '1 o 'fillS B6EURtiMRN'f EUR6N'f2 INS EUR6BRW6RB M2 l'RRtkt eLASSIFlEB B NSA eSSM 123 2 ft YI W ON 1 JtJt4 2818 Declassified and Approved for Release by NSA on '10-'1 2- 20'1 2 pursuant to E O '135 26 vl DR Case # 54778 DOCID 4019683 j -' Published Monthlv by PI Techniques and Standards for the Personnel of Operations VOL VII No 4 - 6 APRIL - JUNE 1980 PUBLISHER WILLIAM LUTWINIAK BOARD OF EDITORS Editor-in-Chief o David H Williams 1103s 1L - Collection Cryptanalysis 1 Cryptolinl uistics J Information Science J ----Jlc8555s IC4902s IfS98 Is Language L Machine support Mathematics 1 30345 IEa Ii IC8518s 1 Special Research Vera R Filby 7119s Traffic Analysis Don Taurone 35735 For individual subscriptions send name and organizational designator to SOg CRYPTOLOG PI P L 86-36 DOClD 4019tiB3 UNCLASSIFIED An address given earlier this year by Hr Phillips before the Computer and Information Science Institute CECIL J PHILLIPS 14 DATA FLOW Challenge of the 1980s1U Jhen I was asked if I would give a talk CISI and I agreed I knew what I wanted to talk about but I was not sure what kind of title I wanted to give it I settled on the Data Flow title although a more proper title might have been Data Flow Feedback Control and Inter-Process Dynamics all of which are involved in what I consider to be our main EDP challenge for the 198Ds U Perhaps I should also state at the outset that the challenge is to everyone in the SIGINT process not just people in the EDP field As we keep saying in our budget justification computers are integral to virtually every step in the SIGINT process either by performing a function by delivering the data to a human to perform a function or by taking the results to the next process at U Let me talk for a minute about data flow's role in the whole picture which as far as I'm concerned is central to the whole process Without data flow there is no need for feedback and control and there is nothing one could call inter-process dynamics U I think people have always understood that knowledge of data flow is useful to understanding processes whether they be human or machine When I first made contact with Automatic Data Processing about 35 years ago one of the basic elements of planning ADP jobs was to produce a'flow chart Flow charts usually showed the movement of material which in our business has always been data-whatever its form In those days flow charts with which I never felt totally at home usually showed functions such as edit punch sort list make corrections etc U These functions were either manual as in the case of edit or manually initiated as in the case of sort list and others The key thing about this is that virtually all the inter-function actions were by humans responding to written or verbal instructions These human interfaces represented both good news and bad news The good news was that function-to-function interface was easy to take care of but the bad news was that it worked only as well as the humans understood and cared about what they were doing U With the introduction of the stored program computer flow charts to show data and control flow became basic tools for planning how to write computer programs usually showing functions at a much more detailed level than the case I described before Stored programs permitted data to be passed from one process to another without human interface but the size of computers and the complexity of programming meant that the data went external to the system time and time again Thus the human interfaces I mentioned were still very much in evidence However there was automatic feedback within programs After all one of the main attributes of a computer was its ability to modify the stored program or modify the data but almost any inter-program feedback was handled by humans U Today there are still hundreds of processes in which inter-process communication is via human beings That would not be so bad if I were talking about humans linking processes through interaction at a CRT But I am not I mean that there are still hundreds of processes where the result goes external to something like a print and any result going to a next process gets entered by hand Sometimes this is via punched cards and there are still a few examples using paper tape to re-enter data U Another of the problems in the way we treat data flow today lies in the fact that too many processes are still batch processes These are probably satisfactory for treating single-direction flows of data but are not very compatible with feedback and controlespecially if we expect feedback and control to have any effect while actions are still taking place I believe there are'much better ways of treating data than these current batch flows and these new methods are also part of the challenge for the 1980s I will get back to this later Apr - Jun 80 Page 1 CRYPTOLOG UNCLASSIFIED DOClD 4019683 CONFIBEN'I'IAL U All of this reference to past and present data flow and flow charting brings me around to what I consider the challenge of the 80s-namely how to make a substantial leap forward in the way the flow of data is understood and treated and to use this knowledge to improve the whole SIGINT process U Why is there a challenge U I believe there is a new kind of challenge'here because I believe that for the first time all the capabilities necessary to cause data to flow from collection to customer with appropriate feedback of control information are present At the same time there are capabilities which will enable analysts-separated by space-to work together as thouRh they were sitting side-by-side I realize that over the years there have been a number of major technological developments each of which seems to be a breakthrough but I believe that the 1980s can see us put it all together well for the first time so as to make a breakthrough in the only area which will have a long-lasting and farreaching effect-namely to begin to process and handle info ation rather than data throughout the SIGINT system E EEe Just for a moment let's take a look at the key ingredients __ Terminals access devices In the 1980s for the first time we will have enough terminals so that virtually everyone in the SIGINT system will have some kind of access to one By the end of the decade we ought to have a terminal or access device on every desk The only weak spot may be in the area of full graphics an area whose application to SIGINT is still to be developed ex ept for some highly specialized cases Concentrators Terminal Sub-Systems We have these in great profusion It is not that these perform a unique function they are just convenient boxes between terminals and mainframes or networks occasionally providing extra computer power or extra storage Networks and Communications These along with the terminals are the real breakthrough For the first time we are beginning to have what is needed-communications to support interactive orocesses at any distance we want I other process via his keyboard U If all of this is true what is the challenge U The challenge is to make it all do some useful work in the production of SIGINT The challenge is to interface people with information not just terminals with computers This is where the payoff the success and the rewards are Since there are a lot of computers already doing many of these things what makes the 1980s special U I believe there are at least three main areas of challenge in the 1980s Causing the data to flow smoothly from collection to output reports Treating the data in natural infor'mation units which foster the development of better analytic approaches and feedback and control Developing techniques for teZeanaZysis a term which I have cointed to describe methods by which two analysts may collaborate over a distance o U Let's take a look at the flow problem first E EE8 Causing SIGINT data to flow effectively and smoothly from collection to user output in natural information units with ap priate feedback and controls sounds like a goal that has been around for a long time but I think that for the first time we can realize the full value of machine interaction and feedback which actually influence processes while they are still going on At a kind of top level design this probably sounds pretty' straightforward At that level it is However at the detail level it means thousands of processes which have to be understood in terms of all the other processes to which they relate These other pro- cesses may be adjacent but thinas which affect them may be several levels away I _ _- -_ _ JI This means that there are so many interrelated effects that understanding them all is out of the question And this is precisely where the greatest chailenge 1 e f fi g e f e W-l 4 c I'-- - l- ia-n-d--by-t-h-e-e-n d-o f-I-9-8-1-t-o- J1c onnect them most major points in the SIGINT systems wher- P L 86-36 U I think it important to understand that what I' am suggesting is not simple-because I am not just talking about standardization at low levels of protocol and format Many of these are already covered in PLATFORM and U In short we are soon to have the technical some of the format conventions are covered in capability for every user analyst at every ter- USSIDs What I am really talking about is a minal to talk to every other analyst and every set of information level protocols-things ever they are Main Computers As you well know there are plenty of these of almost every type Apr - Jun 80 Page 2 CRYPTOLOG CONFI98NTllaL I1ANBbB VIA SShlINIf' 8111ltNNBbS SUbY DOCID 4019683 CONFIDENTIJ I which humans do now in the best cases almost without thinking groups have done a super job often with very limited help from the computer types So there is part of the challenge-to help U Such an example might be where an analyst refine the existing processes and extend knows instantly where a mis-identified message control processes to all parts of the system must go to get it back on track At the same to manage processing analysis and reporting time most of our current processes simply as well as collection It has been said that dump data back into the system The challenge there are more signals in the air than we can of course is to develop automatic re-routing collect more signals collected than we can schemes or man-machine interactive processes process and more signals processed than we which allow for re-entry and re-routing in can transcribe or decrypt and then report on such cases The net effect is that we must be able to select and filter the flow at all points in U I think this dictates the need for dethat flow veloping a convention for annotating and labeling the error transactions so they can U Much of what I have said is fairly simple be re-entered into the process to produce a on a case-by-case basis As I noted earlier new result which is a combination of the comthe complication comes in looking at the puter process and the man-machine interaction whole picture or a large fraction of it process If the person annotating the transThere are also a few other complications in action knows exactly where the erroneous data that broad interprocess communication is not goes the problem is relatively simple If the goal of everyone Systems designers are he simply knows that it is wrong and wants to send it back the problem is much more com- likely to have a goal of optimization within their own domain whtch is usually sub-optiplex This is especially true of the problem mization as far as the overall SIGINT process of how to label it in order to get a different result the second time through the system is concerned Further every manufacturer and creator of new software packages is out there being as creative as possible to give his system some unique features Users of the overall system are likely to have to keep running just to stay even with all the variations U The second part of the challenge had to do with treating data in more natural or event-driven units so that feedback and control have more meaning An alternate way of thinking of this is to consider dealing with units of data Ee eeS Perhaps this is naturally inherent in a good system but we have been so long in orocessinQ modes driven by time or volume U This might be fairly straightforward if all feedback is from a single class of process back to the collector In real life this is not usually the case ' rather the feedback comes from all parts of the system Thus there is the need to be able to accommodate different levels and kindsof information U Up to now we have heen doing all of the things I am discussing-particularly feedback-by human action mostly on a hitor-miss basis There have been some near real time collection control groups but their effectiveness has been limited by slow and erratic feedback to them from analysis by slow and erratic methods for deliv ry of information b lCk to the collectors and often by slow and erratic reporting and communications with the collectors 6 eeS Let me make it clear that these Apr - Jun 80 Page 3 CRYPTOLOG EO 1 4 c P L 86-36 WU QI ii HI' QUI jl' 11 Hr 18bS Q tlbY DOClD 4019683 CONFIBENTIAL U At the same time programmers and systems developers also need to be touched by the twoby-four to remind them that their job must be to simplify the interface to the system so it can be more readily used The challenge to all of you is to help develop the system tools and to help explore the ideas with the analysts who are bold enough to consider them U I think it was Mark Twain who said that there is no death like being talked to death so let me conclude with a brief summation L- -U- ---B-eC-a-u-s-e--t-r-a-n-s-a-c-t-i-o-n-a-l--o-r-'-'e-v-e-n-t---d r i-v e-n d-p-IO-- cessing has strong implications of more timely operations it may seem that I am really discussing time-sensitive processing Transactional processing does make these things possible but it does not demand it You can save work queues in transactional systems as well as in batch and in a transactional system you ought to be able to save them ven more intelligently U A third and final part of the challenge of the 1980s as I see it is develop techniques for teZeanaZysis or teZ syne ism--terms I made up hurriedly to descri e joint analysis done by two or more people over distance Let me explain what I believe sqould be possible here I am talking about a form of teleconferencing in which man-machine interaction takes place with two or m re people at separate terminals operating against the same data where the data may be a essage entries in the data base or any o her form to U There is probably some tendency to equate this to teleconferencing but it is a level above teleconferencing It is teleconferencing in which the analyt1c problem is at least part of the medium U We are optimistic that the seeds of su h ideas can begin develop as soon as network connections are stablished with collaborating analytic fenters The goal is to see the multiple work forces at various centers integrated through this telesynergistic bond so we can use the work force better or tackle more comple i problems U I have no doubt but what it will take a long time to really accomplish this on a broad scale First we have to hit people with twoby-fours to get them to treat problems interactively on computers where one person and one terminal is involved But in time the software and the technology will improve and the probl ms to be solved will become bigger or harder or both so there will be an incentive to evelop telesynergism to U My key point about the challenge is that EDP people and communications people must understand the whole system that is all the parts and all the processes and how they interact with each other whether man or machine Another name for this is total data flow--from start to finish--including feedback U Given some understanding of the system --the depth of understanding will naturally vary with one's role--I see the challenge as one of developing full information interfaces between the parts of the system not just signal and processor interfaces To me this implies overall design and and overall approach--real top down design from the very top not from some local peak U In a larger sense it is the challenge to everyone in SIGINT to make the data flow from process to process and to make the feedback data and contraol data flow where they are needed In short this means to make every last piece of the SIGINT system fully interoperable with all the rest U I think that this is a challenge which can pretty well occupy everyone during the 80s UNCLASSIFIED SOLUTION TO NSA-CROSTIC No 30 From a Letter to the Editor on COPES by Donald Y Barrer PI CRYPTOLOG September 1978 My concern stems from admissions made privately and off the record that when COPES objectives satisfaction is low we alter the objectives so that the ate of satisfaction looks better and when things lopk too good we add objectives to preclude a cut in resources This should not be surprising it is a natural response UNCLASSI FI ED Apr - Jun 80 Page 4 CRYPTOLOG EO 1 4 c P L 86-36 CONFIDENT IIlzPlBbI3 VIA 88MIN'f' enAPUifJM 14L i DOCID ' 4019683 UNCLASSIFIED R TAR FF Rr iRLYST L KS RT fjPUTEAS L -- lp14 n the world of computers I am an amateur I have used them as tools in various analytic jobs for many years but always with someone else acting as intermediary either a programmer or a systems analyst I went through most of the same frustrations and associated withdrawal symptoms that other analytic people did when our traffic was taken out of our hands and file cabinets and put into the bowels of the basement somewhere out of sight and out of reach except through the offices of these intermediaries I I frankly think that contrary to popular opinion we analysts were not all afraid of computers Maybe some were but not all But what we all did share was the realization that our processing cycle was no longer solely under our control Our data was in a loop that went through someone else's area of control while it was in the computer we had no control over it Someone had control of course but that someone was never us The people running the system did not work for the analytic people This is not to say that they did a bad job or that they did not try to support us in the best way they knew But prior to computers when traffic came into my shop it was my traffic and I kept it under my control for as long as I had need of it The people who handled it for me worked for me and were rated by me If that sounds proprietary it is because it was a proprietary kind of system Today however there is very little of the proprietary feeling among analytic people It isn't really my traffic any more All this is not meant as a harangue about computers but is put forward to show where I'm coming from so you can put the remainder of these remarks in context While I have not yet relinquished my sheltered role with respect to computers I have been spending some time at the keyboards of several CRT terminals For the most part I have been trying to do or simulate an analytic task usually some aspect of traffic analysis Since any traffic analyst these days must work within the frame- work of computers all of his data is inside them somewhere I was motivated to see just what doing TA on 'a terminal really amounted to This represents a first 'report of that venture with all of the biases that first reports commonly have I am not finished looking but I thought I would put down three early impressions based on my own hours at the keyboards As I look over my early notes to myself I am impressed about how often I had typed in a TAPQE problem involving several days traffic all in upper case just like an old-fashioned Morse problem for those of you who remember and since all the commands I was using were in lower case I kept finding bits of lower case data in the traffic Later as I became more familiar with the problem I think I could have recited each page of the traffic by the time I was through I began to notice that pieces and lines were missing I really don't know why the erasures and writeovers happened in each instance but often I noticed that the cursor on the screen ran a bit behind me when I was entering data I might stop and glance at the cursor position and not realize that the cursor was running behind There may well have been other basic reasons for the losses and accidental writeovers but the experience left me very wary about typing data in at anything near my normal typing speed--which is not terribly high I also spent some time trying to think out defensive strategies for avoiding the loss of data such as always working on a copy of a file rather than the file itself making a habit of storing off the working file fairly often as frequently as every few lines on material I was anxious to protect The problem needs more attention both by computer people and by system designers since we already lose too much data around here by more classical methods and really don't need new and inventive ways to lose still more I don't have any solutions to offer only a warning that the ease with which data can be altered or erased is scary and will not go away just because we don't want to talk about it At least part of the problem lies in perceiving that it has happened at all most of my errors were not noticed until Apr - Jun 80 Page 5 CRYPTOLOG UNCLASSIFIED P L 86-36 DOCID 4019683 UNCLASSIFIED some later time The next problem I had is a personal one I wear glasses--bifocals In order to read the screen I have to bend my head back so that I can see through the bottom part of the lens where its reading prescription is A whole day of that is a little more than I want to think about When my neck gets sore I have to stop My choice I suppose is to buy another set of glasses just for reading CRT screens Or maybe the government will buy them I have heard people say that when we get CRTs on everybody's desk we can go to a paperless system Don't you believe it Have you ever seen a computer printout Those folks use up more paper than we ever did with yards of stuff preceding the actual print Nevertheless the point I want to make is that I found the screen too SIIIall to do anything like a diagram or a tabular lising of continuities Having experimented with both the PQE problem which was a simple continuity problem similar to the final problem in TA-IOO and with several CRYPTOLOG problems I found that I could not keep track of recoveries in any orderly way when all I had to work with was the screen even one like MYCROFT CARONA which can be moved around and divided into parts But all that changed when I decided to try to work with a combination of printout and screen I used the printout as a basic worksheet making notations on it from time to time then adding them to the screen and generating another printout This will work if the printer is fairly close I don't consider the basement close to the second floor and not so busy that one has to wait more than a few minutes for the output Because of my interest in d ing problems on or through the computer I have been interested in encouraging others to try it also If you have access to a terminal and would like to try it let me know Send me a note to wes at carona if you can access PLATFORM or call me on 3360s As analytic people gain more experience in this area we should find ways to plow that experience back into better support and better systems With this in mind we in Pl4 have set up a file called whydontwe where we record various bits and pieces of reactions to the terminal system network or whatever Not all of it is literate and some of it looks sort of dumb when re-read at some later date But it is what we thought or felt at the time and some of my earlier items formed the basis for some of the above I would encourage you to do the same If you want to see what dumb things we said why don't you have a look if you're a CARONA user at the file u3 pin why whydontwe If you have an idea but don't know who to give it to send it to me If there's enough interest in this sort of thing maybe the editor would be agreeable to a regular feature assuming of course that they're printable ------------------------------------------ITem the Editor CRYPTOLa wi'll be happy to pubZish any printable l'esponses to Wayne's soLicitation and to chuckLe priuzteLy over any which are not P16 Language and Cryptologic Library I ITl2 LibrarianoftheP16LanguageartdCrypt610gic Library wishes to advise readers of the excellent collection of older works on cryptology available there some of which date back several centuries Unfortunately items in this collection are not available to be taken out but must be read on the library premises which i re at FANX Room B3526 For further in format ion I Imay be called on 8873s A partial listing of volumes in this collection is given below Bazeries Etienne Le6 Chi66 e6 SeC4 Vevoite6 Paris Librarie Charpentier et Fasquelle 1901 Booth Williams Stone Some AC4o tic Sig e6 06 F n Bacon Boston Houghton Mifflin Company 1909 Breithaupt Christian The M t 06 VecipheJLing 06 the Science 06 Solv-Lng and Read-Lng SeC4e t lII uting Toge-thl ll 1II-Lth a H t tJM cal Accowt t 06 VaJUo lAe tho 06 SeC4e t lII-Uting in U6e Among t the AncienU and in fo e Recent T-Lme6 Helmstedt Germa y 1737 Breithaupt Christian A H t tJM cal C-UtiCAl and Ve taited V qui l-Ltion Conc ng the VaMo YPe6 SeC4e t 1II-Uting EmploJed by the AncienU M weU M by Tho e06 lAMe Recent V-Lntage Toge-thl ll wUIt an Accowt t 06 the AM 06 Veciphl ll-i ng Printed at Helmstedt Germany 1727 Apr - Jun 80 Page 6 CRYPTOLOG UNCLASSIFIED 06 'P L 86-36 DOCID 4019683 TOP SECRET UMBRA UNCLASSIFIED Cardan Jerome Girolamo 21 Boolu 06 Je lOl le CaJuian a Doctoll 06 MUan on Sub Printed at Bas Ie 1554 Champollion Jean GIl ammaLIl e fgyptienne Paris Typographie de Firmin Didot Freres 1836 Donnelly Ignatius The GJLea t Cllyptog1l llln FllMeL6 Baeon' Ciphell bl the SoCaU e d SMkupeall e P la l4 Chicago R S Peale and Co 1888 Droscher Ernst Die Methoden dell Geheim6 6 en Leipzig K F Koehler 1921 Figl Andreas Sy eme du Chi66Il ie1l e Graz Ve ag von Ulr Masers Buchhandlung 1926 Fiske Gertrude Horsford Studiu bl the Bi- Ciphell 06 Fll MeL6 Baeon Boston John W Luce and Son 1913 Frederici Johannes Balthasar CIlyptogllaphia 011 the AIl t 06 S Wlliting Hamburg 1684 Gallup Elizabeth Wells The Bi-Lit Cyphell 06 S FllaneL6 Baeon Third edition Detroit Howard Publishing Company 1901 Greely A W W4Il VepaIl tm Te egllaph Code Washington Government Printing Office 1906 Gross Hans Handbuch 6U1l U eIl al4 Sy em dell tik Part II sixth revised edition Munich Berlin and Leipzig J Schweitzer Verlag Arthur Sellier 1914 Harvey Henry HlVtvey'd Vangu all d Code New York Code Press of Henry Harvey 1892 Jacobus de Silvestris of Florence Ruled 06 Se WIl iting Rome 1526 Kluber Johann Ludwig CIl yptogll aphik a Manual 06 he AIl t 06 Se WIl iting TUbingen 1895 Lacroix Paul La Cllyptogllaphie Paris Adolphe Delahays Libraire-Editeur 1858 Lange Andre and Soudart f A T4aiti de CIl yp ogllaphie Paris Libraire Felix Alcan 1925 Locard Edmond r e de C tique 2 volumes Lyon Joannes Desvigne et Cie 1935 Loria Gino Le Seienze f Vell'Antiea Gll ecla Hodena Antica Tipografia Soliani 1893 Heng John J VuplLtehu and l etio 06 Conll ad A l u andIl e GeIl aJtd 177B- 17BO Baltimore The Johns Hopkins Press 1939 Meyer H R The Colll7lVlc ial Te le gJtaph Code to Me the Re q emenU 06 the London Te tegllaph Congll ud 06 IB79 New York American Code Co 1880 Plum William R The MUi tivuJ Te tegJtaph VUlLing the Civil WaIL in the Uniled S u 2 vollllll s Chicago Jansen McClurg and Company 1882 Porta Giambattista TUIt Ilty Boolu 06 N Sdenee lei den 1644 Romanini Vesin La Cllyptogll aphie Devoille Paris Typographie Hennuver 1857 Author also listed as Vesin C F Sandler Rickard Chi66e1l Stockholm Wahlstrom and Widstrand 1943 Scotus Johan Maria FoUll Boolu 06 Se ltIJr Ui iIg Naples 1563 Sympson S A Nf I Book 06 Cyplt london John Bowles 1750 U S 8ureau of Navigation Navy Department The In tvuaaUoll4l Code 06 SigIl4l s 601l tht lUe 06 AU Natio Washington Government Printing Office 1875 Also available in revised edition IG94 u S Department of State The Ciphell 06 the VepaIl tm 06 S e Washington Government Printing Office 1876 Valerio P Ve La ClUjptogll aphie Paris Ubrairie Mllitaire De L Baudoin 1893 Velasquez Manuel M Codi go UMV a t de CoMupondencla Se e ta Mexico 1926 UNCLASSIFIED and in a more modern vein U I U IChairmaff6f theBookhrell kers' following publications as essential reference works field of cryptolinguistics If YOJ arei l1 terested of these documents calli Jon 1l03s All Buck Stuart July 1977 151 Forum suggestsJhe for those in the in obtaining any are classified TSC H I- - - - - - - - - - - - - - - - - Apr - un 80 Page 7 CRYPTOLOG OP SECRET U 4RRA EO 1 4 c P L 86-36 -----J IItalian BOokbl'eaking fFBS Nationll 1 Security Agency Technical LIterature Series Monograph No 6 1965 u 86-36 CRYPTOLa o r '--------------------- Ii ---_ _ I U P L P L 86-36 DOCID 4019683 TOP SECRET UMBRia The original paper was written in 1943 while bookbreaking was still being done on the Italian Military and Air Force codes Soon after the Italians surrendered themselves and their codebooks The bookbreakers then had an opportunity to compare their results with the original of PEGASO the Italian title for their latest Air Force code A very interesting account of a specific effort which can also serve as a primer on bookbreaking Swift Katharine L Some Problems and Techniques in Bookbreaking U NSA Teahniaal Jo l Vol XI No 1 Winter 1966 U A concise view of what codes and bookbreaking are all about and some interesting war stories Just the thing to start out on Swift Katharine L Standards and Teahniques of Code Reaonstruction U National Security Agency Technical Literature Series Monograph No 5 1965 s-eeSj Reprint of a 1955 paper S51 126 Designed for code problems of the 1950s which were Icodes of classic configuaration Rich in illustrative anecdotes and strong on first principles Written when card sorters were obsolete but computers were still in their infancy A must EO 1 4 c P L 86-36 Swift Katharine L and Oliver Jean Colleated Artiales on Code Reaonstruction U Cryptanalysis Department National Cryptologic School S-212 802 Revised Edition 1976 e eeSj A collection of shorter articles on assorted topicsoJrelevance to bookbreaking It can be thought of as a supple ent to Standards and Teahniques above which serves to increllseits scope From the Editor You should add to the preceding list the following bookbreakers must which h sjlJst been published I I ITh StrwtW'e 0 f Codes Part I P16 Cryptolinguistic Series No 1 1980 S-221 647 Classia Codes U 6 660 An annotated survey of various codes both U S and foreign used during the first half of this century This work also includes a comprehensive bibliography of in-house and open source publications on bookbreaking UNCLASS IFI ED GEOGRAPHIC TRIVIA Each of the United States has a highest and a lowest point Most everyone knows that the highest high point is in Alaska Mount McKinley 20 320 feet while the lowest low is in California Death Valley --282 feet But do you know which state has the lowest high and which has the highest low Many of the largest cities in the U S are not state capitals In fact the state capital with the largest population according to 1975 Census Bureau estimatesJ is the eleumth-ranking U S city What is it And what are the place and show capitals Old-time trivia players to whom Boston has always been the most populous capital will be dismayed to find out that its 1975 population of 636 725 is fourth largest Answers on Page 15 UNCLASS IFI ED Apr - Jun 80 Page 8 CRYPTOLOG TOP SBEURRB'f UMBRA P L 86-36 DOCID 4019683 UNCLASSIFIED 0H 86 36 l g iilmm u- In a jargon-rich environment like NSA we are used to the acronyms and abbreviations that enable us to express complicated thoughts in a reasonable amount of space either physical or temporal It is inevitable that some ab- breviations will have multiple meanings - For the multi-equivalent championship though I think we should nominate the steadfast multiplier K Onecan envision a question on a future version of the Computer Systems PQE Question a 1000 b 1024 c 512 3 d 4096 e DDT Correct Answer a b c e and probably a The classic abbreviation derived from the Greek prefix kiZO- Often written in lower case You can be sure 1000 is meant when the quantity measured is dol- lars or a standard physical unit b A frequent unit of computer memory whose are almost always powers of two as in 64K words meaning 65536 65K is also used and has the advantage of corresponding to the common use as in a above But the use of 1024 is becoming more common because of its precision 1000 is an approximation while 64K 1024 is exact and the ease of computing the size of memory as it doubles 128K while 131K c The somewhat self contradictory octal 10008 A unit often used by those reading displays or dumps d 1000l I have never actually heard this one used I don t mix much with users of Hexadecimal but I suspect it's out there I guess we should be grate- ful that at least we re not using a number system with a large enough base that might also be a digit in which case it would have value 20 e See a recent Agency organziational chart So the next time you hear a make sure you get onto the speaker's base 0H II Where did everybody go As a result of the recent reorganization which did away with the organi- zation and sent its people off to various other elements has been re ceiving a number of returned subscription copies which had been distributed to addresses 7 So if you used to be in K and would like to continue to receive please call on 1103s or drop a note P1 to let us know your new address Apr - Jun 80 Pagers UNCLASSIFIED ii DOCID 4019683 UNCLASSIFIED NSA-Crostic No 31 word D as seen in Word Z by dhw A and speak each other in passing Tales of a Wayside Inn Longfellow -s4 68 166 255 206 6 wds -rr 179 270 222 103 195 19 B Prudent-sounding state capital C Of maximum fusibility D Hollywood leading lady 1924- active 1950-1959 in such films as Crash 240 ill 12 186 272 226 ill 276 210 3210 40 Landing Donouzn's Brain etc Full name E Harmonious-sounding state capital 3 wds F Mechanism of a firearm that expels the spent cartridge G Cereal grain H Rigorous unadorned grave 1 Romantic-sounding town in Florida seat of Osceola County J Full of plots treacherous K Sibling's daughter L Fate at a turn of the card 4 wds M Secessionist province of N Ethiopia N Women as mates says Thurber should be those who have great constitutional strength and are not O Timid full of fears P Learned Q Building for the storage and maintenance of locomotives 132 235 S'6 214 182 Apr - Jun 80 Page 10 CRYPTOLOG UNCLASSIFIED 82 140 38 229 248 DOCID 4019683 UNCLASSIFIED R Knotted S Toast 3 wds T In pointing out why a fifth of Scotch on the bar was preferable to major brain sur- 86 254 67 93 205 6 102 126 75 4 250 261 SO gery he explained I'd rather have a - - - - - 9 wds 224 152 265 83 262 70 225 112 119 136 44 79 U The property of being not separable into s 268 253 184 142 2462l62M 2 0-1- v The original large-scale-robbery-caper movie 1955 W In Germany units of local government X Hairy rough with bristles Y Sour Z The only film in which Word D co-starred opposite her husband 1959 4 wds T88 3 3 192 267 1'522 6I 171 -2-74 42---gg ZI' Capital city of New Caledonia oo o o oo o oo o o o o o oo o o o 50 68 o T A8 o 9 e 19 A 20 U 21 S 22 Z 23 A 24 N 25 L 26 W 34 F 35 V 36 N 37 T 38 Q 39 8 40 D 41 U 42 Z 43 T 56 Q 57 A 58 U 59 8 60 H 61 73 L 74 Z 75 T 76 An y 78 M 91 o 92 V T 94 T 51 2 3 L6 T 7 oo o o o oo oo o o 52 L 53 E 54 o 55 69 P 70 T7l U 72 1 A 85 H 86 T 87 F 88 E 89 0 A 84 U2 5 o o o o e 4 T 18 1 17 K 90 S 101 e 102 T 103 A 104 E 105 F 106 P 100 E 116 H 117J 165 T 166 A116 136T 137 5 138 G 152 T 1532 1 154 W 150 T 151 5 1611 Ulib ' t 11 U i 182 Q 183 N 184 l 185 E 186 D 187 X 199 K 200 5 201 E 216 U 139 e 140 1171 ZI 172Z 1 173 188 Z 189Z 1 190 I 217 X 218 R 219 D 220H 221 1 222 A 234 L i l35 Q 236Z 1 li37 27 X 12 D 13 E 4 U 15 2 16 Y L 19 N 30 8 31 T 32 D 33 Z o oo oo o o o oo o o o o o o 93 L 44 T 45 Z 62 J 63 864 D 79 T 80 J X 95 i311 251 lbS T 266 267 Z 268 U 269 5 270 A 271 t 272 P A o o 141 K '14i U 143 II '144 1511 K 159 8 11 4 11 r J 0 L 14 t A 11 K U 149 Z 65 N 66 Y 67 T 1111 X 82 Q 83 T 96 1 97 o 98 T 99 Z 114 V 115J 14b 147 D 148 A 149 223 N 224 T 225T 226 D 227 Z 228 V 229Q 230e Apr - Jun 80 Page 11 CRYPTOLOG UNCLASSIFIED i 260 M 261 T 273 A 274 Z 2h M 276 DIU U i4b U i47 T 1i411 262'r 263 A 264 P Ullt U l1I - o o o 1911 T 213 L 214 Q 215 R 212 P 239 X 24U Dli41 t li4i UIl43 1 i44 C i45 J 163 I 164 R l ' A 180 I 181 J 191 T 192 Z 193 V 194 K11 15 A 196 J 197 G f o 130 I 131 R 16u N 161 E 162 H t LI 206 A 207 C 208 J 209 Y 210 D 11 oo o 48 127 N 128 L 129 W 252 D 253 U 254 1' 255 A 256 M 257 e 258 5 259 E 249 5 250 W46 Zl 47 107 E 108 D 109 F 110 A 111 V 112 T 113 1 155 P IS6 X 157 M 202 E 203 G 204 U 205 231 L 232 M 233 11 E 28 D 122 W 123 A 124 E 125 S 126 T 1182 1 119 T 120 L 121 W 13l Q1133 A 134 T 135 0 8 10 dhw DOCID 4019683 UNCLASSIFIED LIME-A OBIO LEEM-A PERU For many years befofe there were Russian-language reference aids the final authority on the pronunciation usage and meaning of any Russian word was the late Juliana M She told you how every Russian word was s pposed to be pronounced how it was supposed to be used and what it was supposed to mean If anyone dared to say but they pronounce it or they use it to mean she would angrily say They the Soviets they have bastardized the language those peasants aren't speaking Russian they're speaking Soviet jargonl But what could you do if there was only one oracle in town You'd have to consult it So people would find themselves asking MIss M to render solomonic decisions asking questions like How do the Russiansnot the Soviets-pronounce it Is it Semipalatlnsk Or is it SemipaHtinsk Miss M would answer Well the name of the city comes from the Russian word pa taJ meaning 'tent ' The name of the city means 'city of seven tents ' Hence the correct pronunciation is SemipaHitinsk Smirk on face of disputant A- I told you so scowl on face of disputant B- I know I've heard Russians-not Soviets say Semipalatlnsk However Miss M continues I used to know a man who came from Semipalatinsk and he used to pronounce it Semipalatfnsk The same kind of fight used to rage around the name Murmansk Everyone in the United States knew the name at the beginning of World War II because it was the northern port with the icefree harbor where before America was officially in the war American convoys of merchant ships delivered all those lend-lease shipments that Soviet historians don't seem to recall that the Soviet Union ever got And everyone used to pronounce it Murmansk So imagine the linguists' surprise to hear that the correct pronunciation of Murmansk is MUrmansk But doubt persists between the Murmansk and MUrmansk factions Finally the bastardizers of the pure Russian language publish a dictionary of pronunciations of personal and place names worldwide for the use of Russian radio and television announcers movie directors and the like Naturally the Murmanskers and the MUrmanskers race one another to the M pages There it is sure enough- MUrmansk But then in a more leisurely moment reading for fun the introductory remarks on How to use the dictionary one of the disputants reads that the pronunciations given in this book are the standard ones for use on nationwide radio and television No indication is given of certain nonstandard or local pronunciations for example the local mispronunciation of Murmansk instead of Mlirmansk That's the problem People don't even know how to pronounce the name of their own hometown From Twelve Language Anecdotes in Search of an Author by Arthur J Salemme retired Apr - Jun 80 Page 12 CRYPTOLOG UNCLASSIFIED --DOC D 4 OL 1 9 - 6 8 3OJ - - - __ SECRET P L 86-36 AI T I this article is classified SECRET IR94 IIANBbE VIA CO olINT C1IANNEI S in entirety I Advanced Identification Techniques AIT provide a means for identifying transmitters by their unique 1 R Cbar ctjrist cs In his article _ _ provldes a brlef summary o t he l-story and present status of AlT as well as a glimpse into its future dhw o n early aspect of AIT is Radio Fingerprinting a technique for examining the characteristics of transmitters to determine unique aspects that will allow these transmitters to be identified when seen on another occasion I A Apr - Jun 80 Page 13 CRYPTOLOG SECRET EO 1 4 c P L 86-36 IIANBbB VIA 8QPtIIN 8I1A-tHJEbS 8Nh'i DOCID 4019683 SBCR T Apr - Jun 80 Page 14 CRYPTOLOG 8 8CR T EO 1 4 c p L _86-- 2Q 1I ' JQbE VIA SQ IIN'I' SIIAl'lNBb8 8N bY - - - - - ------------ --- DOCID J I 4019683 SECRET UNCLASSIFIED Answers to GEOGRAPHIC TRIVIA From Page 8 EO 1 4 c tt Frequent travelers to the Eastern Shore who are convince that Delawa m6 36 be the flattest state are almost ocrrect With a high point of 442 feet near Wilmington Delaware comes in second But first place goes to Florida which soars to a high of 345 feet in Walton County in the western panhandle Incidentally when Washington D C becomes a state it will not displace Florida Washington's altitude high near the intersection of Wisconsin Avenue and River Road is a surprising 410 feet The highest low is in Prowers County Colorado 3350 feet where the Arkansas River enters the state of Kansas ttSurprise The most populous state capital is Indianapolis with 725 077 inhabitants It is followed by Honolulu 705 381 and Phoenix 664 721 The ten largest non-capitals are in order New York Chicago Los Angeles Philadelphia Houston Detroit Baltimore Dallas San Diego and San Antonio Washington D C ranks twelfth These figures are from 1975 Census Bureau estimates preliminary information from the 1980 census shows some changes to this ranking UNCLASSIFIED Apr - Jun 80 Page 15 CRYPTOLOG I SECRET DOCID 4019683 TOP SECRE' l JMBK ' SOVIET This article is classified TOP Ee E'F 1 I11BRA in entirety P L 86-36 1 4 c P L 86-36 Apr - Jun 80 Page 16 CRYPTOLOG ------- '-'---'- ----------- _ _ DOCID 4019683 'fOP SECRET UMBRA -c I Apr - Jun 80 Page 17 CRYPTOLOG TOP SfiEURRBT UMB ' 4 c 86-36 DOCID EO 1 4 c P L 86-36 4019683 'fOP 8BCRB'f UMRRlL Apr - Jun 80 Page 18 CRYPTOLOG TOP SEeItE'f tJMBItA DOCID 4019683 'fOP SHeRET UMHRl UNCLASSIFIED WII 10 DD AbDul ''FANX'' P L -----_-----'IA04 If you didn't know the derivation you could imagine that it was actually PHANXEANS Now isn't that a lot more classy almost Grecian sounding FANXITES my foot Of course since the state of Maryland changed the name of the airport from Friendship to Baltimore-Washington or BWI the term FANX has become obsolete This has bothered me for quite a while and I have tried to come up with a name that is more appropriate Believe me it hasn't been easy You can't do much with BWI to make a pronounceable acronym I rej ected Airport Annex since it doesn't tell which airport National Dulles You could go the way of some area wags who tried some time ago to popularize the term Baltington to describe the megalopolis that surrounds the B - W Parkway Personally though Baltington Annex has about as much appeal as FANXITES How about Parkway Annex No it sounds too much like a tacky motel FAN-ex This is somewhat parallel to the difI tried picking out another prominent landference of opinion on how to pronounce the mark in the area The biggest thing around term DIRNSA In this case my informal poll after the airport is the Westinghouse plant shows that a decided majority of Agency emAfter experimenting with the firm's name and ployees prefers the two-syllable DIRN-sa But abbreviation the best I came up with was there is a stubborn minority that insists th Compound W Somehow that wasn't quite it it is a three-syllable word dir-EN-sa But I had almost given up when I stumbled I digress onto the solution The annex is located in I hope that members of both FANX prothe South Linthicum area So what would be nunciation camps were as disturbed as I was more fitting that South Linthicum Annex several years ago when the Credit Union put Accurate and descriptive but not very exout a flyer aimed at those of us who were citing you say Not to worry Like Friendship Annex it is bound to become more popuworking at the airport installation This larly known by its contraction SOLINEX flyer hailed us as Dear FANXITES The cash award I expect to receive from FANXITES the Suggestion Program for this contribution The word made us sound like some kind of will be donated to the Civilian Welfare Fund insects that come out of the woodwork Didn't The satisfaction of having solved the FANX they know what we preferred the more elegant problem will be reward enough for me term FANXEANS ack in the sixties a coworker who lived in Linthicum used to drive to Ft Meade via Elkridge Langing Road near Friendship Airport Along that route construction had begun on a site described as the future location of the Baltimore-Washington Science - Industry Center Very interesting he thought Maybe in a few years I can bring my kids down here to tour some of these scientific and industiral institutions How surprised he was when t was later announced that the occupants of the site which was to become known as Friendship Annex or FANX were to be part of NSA's growing population including him From the very beginning the acronym FANX caused a pronunciation problem My own unscientific investigation reveals about a fifty-fifty split among Agency employees One half is convinced that FANX is a onesyllable word pronounced FANX while the other half composed of equally intelligent people pronounces it with two syllables-- B 1 1 UNCLASSIFIED Apr - Jun 80 Page 19 CRYPTOLOG 'fOP SHeRE'f UMBRA EO 1 4 c P L 86-36 86-36 DOCIO 4019683 CONFIIl8NTIAI HELP MANTEO NOT ICE CHALLENGE NOT ICE Tom Engle Senior Linguist in A64 with his tongue only part way into his cheek offers this prospective recruiting flyer to any Agency element which might wish to use it u NOT ICE NOT ICE ADVANCEMENT EXCITEMENT All of this can be YOURS with a career in A6 l NOT ICE FUTURE Apply todayl IF YOU HAVE A DEGREE IN Aeronautics Chemistry Engineering Finance Geography Geology History Law Mathematics Medicine Physics Science IF YOU HAVE BEEN EMPLOYED AS Cartographer Communications Specialist Computer Specialish Mechanic Newsman Broadcaster Photographer Political Analyst Politician Test Pilot Truck Driver Weather Forecaster Xray Technician OR IF YOU HAVE Been an admiral Been an astronout Been a general BuH t a rocket Engaged in international trade Launched a satellite Managed an industrial enterprise Negotiated an arms agreement Planned military operations Run a rail road Served as a diplomat Served on a Military Advisory Group APPLY TODAY DON'T DELAY POSITION OFFERED QUALIFICATIONS 5 10 20 ALL of of of of above above above above NOTE SUbjects subjects subjects subjects GG-S 7 TRANSCRIBER TRAINEE GG-9 ll JOURNEYMAN TRANSC IBER GG-12 SENIOR TRANSCRIBER GG-13 SENIOR LINGUIST ALL APPLICANTS MUST BE QUALIFIED RUSSIAN LINGUISTS Apr - Jun 80 Page 20 CRYPTOLOG CONFI g HTIAI -------- PI-Nov 81-8 -8738 $4 mwmm ow 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