x a2 NSA on 10-17-2012 ursuantto E O 13526 MDR Case # 5477 D ' 1300 Vtf alWW lBl l i lBl lBlI 1lIl i l l f llI i n UiIlll lIIaltfVlDliiaik lTl I 1st Issue 1987 1 4 c P L 86-36 P16 fr L BOW TO WRITE A MEMO U 86-36 t CHECKLIST FOR NON-PROFESSIONAL INTERPRETERS q o o Obst GOLDEN OLDIE U HOW DID A NICE CRYPPIE LIKE BULLETIN BOARD U FROM THE PAST U CONFERENCE REPORT U 'NSA' PUZZLE U YOU U a d van Reigersberg 14 ' 1 o o 1 16 17 19 20 21 29 I I IIANOLE ' IIA COMINT CHANN LS ONLY CLASSIFIED BY NSA CSSM 123-2 DECLASSIFY ON OrigiRatiRg NOT RELEASABLE TO CONTRACTORS AgeREY's eterffliAatieA ReEitlife S CHT- I DOCID 4010012 Fr' tL 86-36 PubllshE d by PI Techniques and Standards ' ' ' Vb bXIV No 1 PUB A VALENTINE to 1st Issue 1987 '-- LiSFfR ---' THE FRIENDS OF CRYPTOLOG BOAR QF EDITORS Editor 'C Collection Computer Systems ' ' 1 963-1103 1 1 ii tiC S k963-5877 1 963-1103 o o ' ' o o' ' o o' s 963-5292 Index o ' ' ' Information Science 963-3456 Information Security o George F Jelen 859-1211 b 963-3845 Intelligence Research o 1 1 1 963-3057 Language Mathematics 1 963-5566 Puzzles 963-6430 Science and Technology 1 968-8075 Vera R Filby 968-8014 Special Research Traffic Analysis Robert J Hanyok 963-5734 Illustrators 11 - 1 l1963-3057 1963-6211 To submit articles or letters by mail send to Editor CRYPTOLOG PI HQ 8A18 If you used a word processor please include the mag card floppy or diskette along with your hard copy with a notation as to what equipment operating system and software you used via PLATFORM mail send to NOTE CHANGE r y p t 19 at barlc05 bar-one-c-zero-five note no '0' Many people behind the scenes make it possible to publish CRYPrOLOG They are in the Press and in Distribution they are on the Printing Control Staff they do data conversion they review for These people carry out their classification assigned duties with extra-special TLC To each of them we send a valentine We are also most indebted to our good neighbors in N the editorial staff of the Cryptologic Quarterly and the system administrators in P044 and in '1'09 for their bountiful professional courtesy They very kindly allow us to use their printers while ours is in storage and rescue us during our frequent crashes To each of them we send a valentine And we are grateful to our colleagues in PI They are the invisible home team who do odd jobs as needed To each of them we send a valentine And last but not least we cherish the many friends among the readers who steer us to the hot items of the day To each of them we send a valentine Always include your full name organization and building room and secure phone number For Change of Address mail name and old and new organi zations to Editor CRYPTOLOG PI HQ 8A 187 Please do not phone for Contents of CRYPTOLOG should not be reproduced or disseminated outside the National Security Agency without the permission of the Publisher Inquiries regarding reproduction and dissemination should be directed to the Editor o'OR OI'TICIAL US ONLY CRYPrOLOG ----------- I EO 1 4 c P L 86-36 seCReCf P L 86-36 1st Issue 1987 o CRYPTOLOG o page 1 SeCRSCf IIiHlBLB YIlt eOl HP'ofCf eIIA H'H3LS EO 1 4 c ft 86-36 Op d DOCID 4010012 EO 1 4 c P L 86-36 1st Issue 1987 CRYPTOLOG page 2 SECRE'f IIAtWLEl VIA GOMUt' ' GIIAUNElLS OULY DOCID 4010012 Sfi3CRfi3't' EO 1 4 c F L 86-36 1st Issue 1987 CRYPTOLOG page 3 ttA f LE Sfi3 CRfi3'f' '11 e6MIft 't' eHoAU t ELS OIt LY I DOCID 4010012 SElEURREl'-F 1st Issue 1987 EO 1 4 c P L 86-36 CRYPTOLOG page 4 SIWIlE't' HANDLEl VIA COMIN'f CHANUELS ONLY DOCID 4010012 SKcaKT 1st Issue 1987 IIA mLB EO 1 4 c F L 86-36 CRYPTOLOG page 5 SB6RS'P YUt eOr HU'f' CIIAUNELS ONLY I r DOCID P L 86-36 4010012 CRYPTOLOG page 6 SEEURREq ' Q fA COMU 't' CIL'UHIKbS OW1 1st Issue 1987 flANnL L DOCID 4010012 SBCRB'f 1st Issue 1987 IIA mbE CRYPTOLOG page 7 SF36RI'3'f VIA eOMTN'P CIIAPHUlLS QP LY 86-36 DOCID I 4010012 EO 1 4 c P L 86-36 1st Issue 1987 CRYPTOLOG page 8 SIi3CRE'P HAUDLE VIA eOM-Hi'T 61h H'lKb8 O tbY I DOCID 4010012 P L UNITED STATES GOVERNMENT 86-36 memorandum DATE REPLY TO ATTN OF SUBJECT TO March 1987 SERIAL SC L - HOW TO WRITE A MEMO All Memo Writers 1 The purpose of this memo is to provide a set of simple guidelines for writing a memo My goal is to improve office communications by making memos easier to write and easier to read 2 The reason for doing this is that I find memo writing in this Agency to be in a deplorable state problem Memo structure has long been a particular Trying to find out why the writer sent me a memo and what I am expected to do about it is like solving a whodunit Grammar is another but poor grammar is so widespread that I try to overlook it Lately memos have become rude Request They seem to command me where a clearer I request or a polite please would have warmed me to the task of reading the memo and trying to do what the reader wanted Somehow I have managed to get along with all these problems But the situation has reached a crisis not even communicate say Now I find that many memos do I no longer understand what they are trying to The time has come to do something about it Hence this article 3 My guidelines are Present your information in the prescribed order Write for the person who is going to read it Be brief Use attachments for details Use words correctly and write simply and clearly Be polite and courteous Be grammatical 4 A more detailed discussion of these Guidelines is attached 5 The action to take is this please follow these guidelines from now on whenever you write a memo 1st Issue 1987 CRYPTOLOG page 9 FOR OFPICIAL USfoJ ONLY I DOCID 4010012 ATTACHMENT A memo is among other things usually a brief communication that contains directive advisory or informative matter So says Mr Webster certainly correct so far as NSA is concerned And he is we use memos in this Agency mainly to advise inform and direct Guideline 1 Present your information in the prescribed order Memo structure is designed to assist this process of telling the reader something in the most succinct way possible serves a purpose Hence every paragraph in a memo Look back at the memo in the beginning of this article and notice that I have underlined certain beginning words in three of the numbered sections memo These words provide a formula for structuring the contents of a You can't go far wrong if you stick pretty closely to this simple format The opening paragraph of every memo should clearly tell the reader why you have sent the memo and what action if any you want taken that Ben Franklin gave to newspaper reporters applies here what when paragraph first paragraph reader The advice this is the who Always without exception put this information in the This quickly places the rest of the memo into context for the one reads on with an understanding of what it is one is reading why one is reading it and what action one must take Do not bury what you want done in the text where the reader has to hunt for it or depend on the summary in the last paragraph to do the job The next one or two paragraphs convey pertinent information to the reader Depending on the kind of memo they may be brief background or history concise explanatory detail to define or describe the subject as a reminder or because it's complex the reasons why you have done something like The reason for doing this is paragraph in my opening memo the authorities under which you are asking that something be done or the specific information for which the memo is being sent such as the Guidelines in paragraph three of my memo above o The final paragraph summarizes the action you expect the reader to take and gives instructions for complying including the expected due date a point of contact for questions and a telephone number If you ever wonder why people don't respond to your memos as requested perhaps it's because you have hidden this information rather than highlighted it Guideline 2 Write for the person who is going to read it I'm tired of getting memos that read as if they were written by a computer-stilted wooden convoluted and often comprehensible only to another machine It is easy to write a memo if you write as if you are talking directly to 1st Issue 198 o CRYPTOLOG o page 10 t'Ol OFFlClAL LJSI ONLY I DOCID 4010012 someone The technique works even when you are writing a memo destined for numerous readers Here are some examples to show what I mean A Dick 1 This has happened and I'd like you to do so and so about it please This one has a short fuse so I'd appreciate hearing from you by 2 Here's how it all came about o 3 Please get back to me on this by o Call my Exec on o if you have any questions Thanks very much for your help B Mary 1 You asked me this question o repeat the question o and here is your answer o 2 I got the information by talking to so and so and by phoning some contacts I have at such and such with the answer I have given you Everyone or whoever agrees 3 If you have any questions about this or want me to pursue it further please just call me on o o C Jim 1 We've recently established a new procedure or committee or organization and I want to tell you about it You don't have to do anything just read on 2 The reason we have done this is that we saw a need for a new approach to such and such 3 The officer in charge of this is o o o on extension o o She will be glad to answer any questions you may have I hope that this new procedure or committee or organization will operate to our mutual benefit and satisfaction D Anne 1 We've recently finished our draft of the PPP Report for the Director and we are now requesting your help in reviewing it If you 1st Issue 1987 CRYPTOLOG page 11 FOR OFI'ICIAL USI ONL DOCID 4 lQR your I comments to us by o o we'll be able to meet our deadline of oo o for submitting the report to the Director 2 You'll remember that this project which is designed to o oooo started in ooooooooo 3 Thanks for your quick attention to this request Officer My Action oooo o will be glad to answer any questions you may have on oooooo Please send your comments to him by oo o ooo o Guideline 3 Be brief Use attachments for details If you follow the structure prescribed above and say what you have to say succinctly then your memos will be brief and they will also do what you intend them to do--communicate One of the reasons memos have become so difficult to understand in addition to being so disorganized is that they are often far too long and rambling The reader gets lost in all the unnecessary details if he or she doesn't get bored first and give up reading altogether Details do have an important place in memos but that place is in the Attachments Use the memo text to direct the reader to what is included in the Attachments Elaborating text charts displays copies of pertinent laws referenced documents -- these all belong in the Attachments clearly tabbed and indexed for ease in referring to them In the memo itself stick to what you really want to tell the reader without any distracting information Guideline 4 Use words correctly and write simply and clearly Memo writers often succumb to the temptation of trying to impress the reader The irony is often that the writer displays ignorance instead by using words incorrectly impressive Enormity is a good example A thing can hardly be bigger hugeness it means wickedness outrage or crime ' me you enorm i ty of this project overwhelms crooked that I can't stand it It sounds so important and But enormity doesn t mean When you exclaim that The are saying This project is so Viable is another good example really know anymore what it means Does anyone Popular usage is eroding nice distinctions and making communication more difficult in the process Even if you use a fancy word properly it is not necessarily in good memo writing taste or style to use it Aim for elegant simplicity which is always more desirable in a memo--and usually safer will send your reader to the dictionary Don't ever write anything that Always remember that you are trying to tell your reader something important and you want to make sure it is understood quickly and perfectly simply Brief memos help you to write clearly and They don't leave much room for lengthy words or text I also recommend an informal style for most memos as you can see from my examples above Don't be afraid to use informality when it is suitable An informal style makes it easier to write simply and directly and to avoid the 1st Issue 1987 CRYPTOLOG page 12 f'On OFI'IGLt USIC ONL DOCID 4010012 stuffiness many memos have or the appearance that they-were written by a computer Guideline 5 Be polite and courteous Whatever has happened to the little word please Nowadays memos all seem to issue commands the 1II0st popular of which is Request just the bald command Request but it is no longer desirable That's all I realize this was once an acceptable usage It is jarring to the modern ear At the very least you should say 1 request so and so or urge suggest ask etc But even better is Please do such and such or I would appreciate it if you would do such and so for YOU a favor After all you are asking the reader to do something You should show that you realize this and are grateful Ordering a person around is not the best way to get something done for you A little courtesy goes a long way Guideline 6 Be grammatical Poor grammar and word usage often interfere seriously with the reader's ability to understand memos Good grammar is a primary asset in communication and should always be striven for More on grammar another time Author's Postcript If you think it's a chore to write memos I think it's a chore to read most of them Memos these days tend to start in media res They lead off with what should be the middle paragraphs omitting entirely any introduction to the subject to explain why the memo was sent to me and what I am expected to do as a result For years I have been reading the la t paragraphs of memos first in order to find out If you don't believe me test it out right now qn the memos in your box Reading a memo in the order written is like reading a mystery story in which the whole thing doesn't come clear until the very end That is good strategy for mystery story writers but not for memo writers While you're at it you might count how many memos there are in your box Busy executives around here all have mailboxes stuffed with memos and only limited time to read them On a typical day the Chief of Staff plows through some eight to ten inches of them The Deputy Directors have equally stunning loads And all other managers and supervisors get their share Keep in mind that you are writing a memo to communicate something you think is important for the receiver to read and that your reader is probably a very busy person In such circumstances you've got to do everything you can to make sure you get your message across for your readers sake and for your own Editors Note An example of how not to write a memo can be found on page 20 Yes it's for real' The indicative information was removed 1st Issue 1987 CRYPTOLOG page 13 FOR OFFICIAL USlt ONLY I DOCID 4010012 o CHECKLIST o FOR o NON-PROFESSIONAL o INTERPRETERS o U by Harry Obst Director Director Office of Language Services and Stephanie van Reigersberg Chief Interpreting Division DEPARTMENT Editor's Note This article may not be reprinted in another publication without the express permission of Mr Obst of STAn 2 Make yourself aware of the circumstances surrounding the meeting Read up on the issues to be dealt with Ask for the minutes of prior meetings if available This list was prepared by the Office of Language 3 Read briefing materials and talking points Services to provide some guidance to Foreign carefully with a view to saying their contents in Service Officers and others who may be called upon the other language Be sure you know what all the to interpret for high-ranking officials when no acronyms mean and how to say them in both professional interpreters are available languages I Pre-meeting Preparations 4 Do not underestimate the difficulty of rendering some simple sounding English phrases and expressions in other languages Try to say I am looking forward to working with you or we hope to be forthcoming in the language into which you will be interpreting for example Likewise do work on rendering the exact psychological emotional charge of expressions in context not just the dictionary equivalent 1 Do not agree to interpret at a meeting about which you have not been briefed in advance or if the subject matter is totally unfamiliar to you Make sure in which direction you are supposed to interpret You may be able to handle Russian into English but not English into Russian Make sure that you are not expected to do simultaneous 5 Try to obtain and read the biodata on the interpreting involving equipment and usually participants before the meeting teams of two or more interpreters unless you have had training and experience in it This is best left 6 Speak to the person arranging the physical to professionals and even some of them can handle set-up for the meeting and arrange to sit in the only consecutive interpreting proper place In general there are expections this means the following 1st Issue 1987 CRYPTOLOG page 14 POll OFFICIAL US ONLY I DOCID 4010012 a next to your principal if you are interpreting in the other language of what he is saying In for him alone short do not summarize or censor 4 Do not draw out the interpretation of the social niceties at the opening of the conversation Move through this at a faster clip as the principals c to the principal's left if you are at a are already losing much of their allotted time due conference table or if the meeting is a meal to the need for interpretation Conversely make sure to slow down a little and to apply the 7 Find out in advance how many persons need interpreting assistance Do not assume it is the necessary emphasis when you interpret an important point or come to the crux of the message principal alone just because you were asked to so the principal will not miss it interpretfor him or her b between the two principals if you are to interpret both ways 8 Find a moment to ask the principal directly or if necessary ask one of his stam how he wishes to proceed If he asks for full consecutive or whispered simultaneous interpretation and you feel that you are not able to perform well in one of those modes tell him before the meeting starts If you are uncomfortable in both techniques ask him to go a sentence or so at a time 5 If the speaker talks too long for your skill level politely remind him that he should pause for interpretation Do not be embarrassed about this because you are there to provide a service a communications link not to put on a bravado performance II Conduct During the Meeting 8 If you are called upon to interpret a toast you may write it down if you don't have the text in advance where you are seated provided you can hear clearly However at least while you deliver it stand with your principal and do your utmost to project the same atmospherics as he does cordiality friendliness etc Be sure to look up at the guests occasionally 6 Stay close to the principal during stand-up and walk-around receptions and similar situations 9 Make sure you have two pens and a note pad The principal must be able to summon you in 5 with you at all times Use a small note pad at seconds Do not get involved in you own socializing public ceremonies formal dinners etc A legal size and let him drift away pad is not easily carried in the shuffie security 7 Pace yourself Don't give everything you've reporters waiters and it spoils the pictures taken got in the first 10 minutes -- you may have to keep by the media and the official photographers at it for an hour or more 1 The golden rule for the placement of interpreters is simple Sit where you can clearly hear what needs to be interpreted Do not be shy about moving if you are shunted off to a spot where the speakers are inaudible Speak clearly and with self-confidence loud enough to be heard by everybody but not louder than necessary At the end of the luncheon or dinner 2 Speak in the first person Never say He says to thank you for coming or use any form of speech when you come to the toast itself it is advisable to depart from the rule that the indirect discourse interpreter always speaks in the first person If the 3 Avoid interjecting any personal opinion of guests have already drunk the toast because so the he says but probably means variety As a many of them understood the speaker and are now substantive officer you may worry when your seated listening to the interpretation switch to the principal strays from his talking points but as an third person and past tense at the very end and interpreter your role is to present a mirror image say for instance and then the Secretary raised 1st Issue 1987 o CRYPTOLOG o page 15 FOR OITI61AL USE ONLY I DOCID 4010012 CONFIQ8NTIAL his glass and drank a toast to the health of the President and to the continued prosperity of EO 1 4 c P L 86-36 Golden Oldie III After the Meeting 1 In most meeting there is a notetaker who will write a memorandum of conversation If in the course of an interpreting mission you found yourself briefly alone with the principals with no notetaker present and something of substance or importance being communicated go to the principal's chief aide after the meeting and volunteer to write a memorandum of conversation on that exchange He will instruct you how to write such a memo or may be satisfied with an oral debriefing This obviously does not apply if your principal instructed you to make no record 2 Return all classified briefing material and destroy your notes if no memcon is needed Do not retain any copies of the memorandums of conversation Destroy or surrender your interpreting notes after the memorandum is finished unless you are instructed to hold them until after the memcon is reviewed and accepted 0 Solution to NSA -Crostic #63 plus December 1986 Lambros D Callimahos Ars Conjectandi The Fundamentafs of Cryptodiagnosis DEFINITION diagnostician n An experienced cryptanalyst of ability just before retirement age This is but a gentle reminder to the reader of an aphorism of Hippocrates as translated by Chaucer The Iyf so short the craft so long to Ierne PLUS The Bonus or Clue as the case may be is in the fifth position down and reads e6NI IElI' q'fIAL IIANBbl3 YIA SOMUil' SIIMiK iJbS FIND HIDDEN MESSAGE CALL DLP 1st Issue o CRYPTOLOG o page 16 eONIi'IBSN'PIAL fl1tf q jLE VIA COMfU'f' CHAPHiELS OP'IL O fbY DOCID 4010012 eONFIB8N't'IAL WHAT'S A NICE CRYPPIE LIKE YOU DOING IN A PLACE LIKE N L uJ 86-36 r-------------I A509 CONFIOEPHIAl in its entirety Quitting time April 1985 I was on my way out the door when the secure telephone rang Ms Wilson GALAXY has surfaced your name as a top candidate for a position in M3 M3 Personnel What does Personnel want with a cryptanalyst N3 not M3 Payroll is going to encrypt paychecks Curiosity being inherent in cryptanalysts I agreed to an interview and began one of the hardest fun jobs I've had in 18t years at NSA look at the planning and to recommend approval of the money and actions AWESOME Yet my power was not inordinate I really could only recommend a position to the Deputy Director for Programs and Resources DDPR through a chain of Division Chief Office Chief and Group Chief That kept me humble I was often grateful that my view as the sole cryppie in DDPR was tempere'd by the perspectives of the very experienced managers in the N3 chain All the while the time clock was ticking away at my tour in DDPR Two years in N342 was considered optimal after which I was expected to go back into the DDO mainstream In the meantime I would be part of a select group in N342 known as System Management Officers or SMOs each of whom was a resident expert for a different discipline signals analysis language analysis and voice processing telecommunications collection computer And what was this hardest fun job Very science intelligence research engineering and simply to look at from an overall Agency point of course cryptanalysis After two years out of of view all of the projects in support of the mainstream the theory is SMOs lose touch cr tanal sis with what is happening in the trenches in their field and fresh blood is needed to provide a current perspective Also the period of two years allows the SMO to participate in at least 1- got to L two budget cycles and to learn how the A certified cryptanalyst I had never worked outside DDO worse yet except for my intern tours I knew little outside the workings of A5 Oh there were occasional conversations with G4 and W3 But with regard to the mission of the Agency overall I was a cloistered nun dedicated to the worship of bits ignorant of any larger world outside my cell ' ---------' P L 86-36 1st Issue 1987 CRYPTOLOG pdge CQNFIQBN't'IAL 17 DOCID 4010012 F L 86-36 GONFIBBN' IAL Agency gets money this is an excellent ticket for managerial positions As the resident cryptanalyst in DDPR nominally my job was twofold 1 to enforce NSA Circular 25-5 and 2 to advise the DDPR on cryptanalytic programs from a macro overall NSA perspective The DDPR as the Agency Directorate for Programs and Resources is responsible for making sure that NSA acquires needed resources and spends money responsibly And we must account to three layers of oversight Department of Defense Director of Central Intelligence and always Congress A primary mechanism for insuring adequate planning and senior review approval is NSA Circular 25-5 That blankety-blank 25-5 It calls for documentation of all NSA initiatives projects costing $2 million or more with coordination and review by appropriate Agency-level authorities at specific decision points in the acquisition - - - - - - - - - - - - - - - - - - - - - - These necessary coordinations allow L to plan space T to plan comms circuits E to gear up for new training courses And as the details are being worked out it gives you a chance to clarify the points that the en neer misunderstood Finally the engineer writes a plan laying out the details of how to get your processor assigning responsibility to any OPI that will be involved in getting your system and showing a schedule of milestones for keeping the acquisition on time This is a System Acquisition Plan and agarn you and appropriate offices in L T E etc will coordinate and have a last chance to make it say what you want it to do Once it's coordinated it is the record documentation for your project or acquisition and the engineer can now put out Procurement Requests PRs to buy whatever is needed As DDPR's coordination staff for 25-5 documentation the N34 SMOs review all of these papers DDPR is one of the few places where all of NSA's efforts on a specific problem can be viewed for both serious holes and wasteful duplication The budget resource managers in N2 understand the figures and totals and how to balance the overall NSA budget They do not however have the technical expertise to understand the actual impact of budget cuts on the people in the trenches and on the work itself The perspectives of the N34 SMOs Who are evaluating the efforts and who can see the real effect of a projected cut from the point of view of the working cryptanalyst signals analyst etc enable the DDPR to operate with better understanding The biggest problem I had was my desire to save every project that cam e across my desk because each one addressed what I considered to be a valid cryptanalytic need Sometimes I succeeded in rescuing a program from the budget cutters or from shooting itself in the foot because of poorly written justification While I was sometimes likened to a terrorist - - - - - - - - - - - - - - - - - - - - for enforcing 25-5 and related controls just as P L 8 6- 3 6 often I was thanked by the sponsor the user 1st Issue 1987 CRYPTOLOG page GONFIBBNtflAL 18 EO 1 4 c P L 86-36 DOCID 4010012 C6NFIBRN't'IAL or even the developer as being the policeman on the beat who came to their rescue when a project was being hijacked in political power plays or threatened by budget cutters or abandoned because of resource crunches Behind-the-scenes liaison with senior staffs was common and I really came to understand the organization of NSA for the first time in my career Visibility was tremendous for better or worse I briefed DIRNSA as well as the Deputy Directors on an aspect of Advanced Modulations and I briefed the Assistant Director for Installation and Logistics on an anticipated avalanche of future projects And senior managers in Rand T places I never even knew existed when I was in A544 bandied my name about I also realized that cryptanalysis was in much better shape than I would have guessed from my futile attempts when I was in A544 to get money to upgrade some obsolete collection positions A hard job You bet The responsibility for reviewing and evaluating cryptanalytic projects across the board is serious and mentally draining But definitely a fun job For it means that you get to see from start to finish everything the Agency is doing in your field and have a chance to comment on it What a tremendous opportunity And that's why a nice cryppie or signals analyst or linguist etc should consider going to a place like N 0 BULLETIN BOARD OLD PHOTOS WANTED U U Do you have any old photographs that illustrate an aspect of Agency history A picture of an intercept site Photos of obsolete R D projects Candid pictures of NSAers at work Formal ground-breaking or VIP visits U Whether they are organizational pictures incorporated in an old briefing or personal photographs stored in your desk they might of interest to the History Program NSA now has han official archive of historical pictures maintained byT54 U If you do have some pictures that illustrate an aspect of our history and you can identify the subject and date with reasonable accuracy please let us look at them If you need the originals back we will make a copy for the Photo Archives U We also accept donations of photos and s l ides Call or positives and negatives _ _ _----'l972 2355 J ' L 86- 3 6 L I FOR CRYPTOMATHEMATICIANS FOUO NSA has a resource which every cryptomathemician should know and use It is the R51 classified mathematics library located at FANX-III Room B2B37 and presided over by librarian This collection of papers contains some real gems of mathematical cryptology dating from the years of World War II as well as much significant research of more recent times His well worth a trip to discover this reservoir of interesting and useful material Mrs ___ an be reached on 968-8580s I I CRYPT PROBLEMS WANTED FOUO KRYPTOS and the NCS will be publishing a series of problems in cryptanalytic diagnosis that will be representative of current target systems The intent is to publish the problem in one issue and the annotated solution in the subsequent one It will be designed so that the set of solutions can be bound and thus collectively form a diagnostician's desk reference FOR GERMAN LINGUISTS U Taped German radio programs with transcription glossary and extensive cultural notes can be loaned to interested individuals The program is Schau ins Land produced 14 times during the academic year Write to Ip16 HQ 8A187 I FOUO We request your assistance in finding suitable problems They should be non-trivial but capable of at least a partial diagnosis on the basis of a small sample and or a statistical summary If you have such problems classified no higher than TSC please check with your management first then get in touch wit r15 963-3957 CRYPTOLOG page 19 COfU IBliH 't'IAL 1st Issue 1987 P L 86-36 P L 86-36 DOCID 4010012 FROM THE PAST UNITED STATES GOVERNMENT memorandum SERIAL Xxx xxx 72 4 May 1972 x Xxxxxxx xxxx xxx SUBJECT TO REFERENCES XXXXX Xxxxx Contact Report i ng Xxxx Xxxx-xxx Xxx Xxx a XXXX-3-100 4 Apr 72 b x xxxx xxxx x 27 Mar 72 c Xxx-066-72 4 Mar 72 d Xxx-065-72 4 Mar 72 1 Inclosed for your action is a copy of references a and b which respond to references c and d A copy of references d e and f clted in reference b above has been also inclosed for your convenience Two copies of reference c XXXX XXX-Xx cited in reference b were forwarded to Xxxx by XXX on 19 February 1971 The delay in forwarding this material was caused by the necessity to obtain a copy of reference f cited in reference b above from XXXXX 2 Please forward your comments on reference b above t Xxx so that a response may be prepared and sent to XXXX-X as requested in reference a lsI Xxxxx X Xxxxx for Xxx X Xxxxxxxx Chief Xxx Incls a s cc Xxx less incls Xxxx Xxx Xxx refs a and b only 1st Issue 1987 CRYPTOLOG page FOR OfT ClltL USS 20 O LY DOCID 4010012 Conference Report Additional information and some of the papers can be obtained from the attendees The 1987 Joint Mathematics Meetings San Antonio Texas 20-26 January 1987 BENEFITS OF ATTENDING Reported by David Harris R513 This is the national meeting of the two major mathematics societies the American Mathematical Society AMS and the Mathematical Association of America MAA along with some other organizations It drew some 3500 attendees an t e aut or orne 0 us concentrate on talking with colleagues making contacts and studying recent developments in mathematics and statistics Others were engaged in formal recruiting It is important that the Agency continue to be represented at such conferences It provides us with an opportunity to talk with outside mathematicians and to follow developments and thus enables us to monitor the mathematics done in the public sector Conferences also give us an opportunity to improve public relations DoD continues to have public relations difficulties with the mathematics community and this has affected our recent efforts to establish a funding program for outside mathematical work By attending these conferences we are in a better position to anticipate what the problems will be and to develop strategies for meeting them Also in attendance were individuals who are associated with NSA or with IDA CRD Much of the benefit from going to conferences derives from informal talks with mathematicians as we had with Thurston Calderbank Diaconis Kailath Nevai Askey Among the other attendees were notables in and Lax detailed below We get to be considered as human beings and so become less mathematics from industry and academia threatening to the academics Through Th e following is a summary of my own informal talks we learn more readily what we view of the major points of interest at the may expect in the way of technical breakconference For reasons of space only a few of throughs While we un the risk of giving the mathematical papers are summarized here academics information on our capabilities we P L 86-36 lstlssue 1987 paR CRYPTOLOG page OPI IEURIAI 21 use ONLY DOCID 4010012 also get to judge how close we are to the outside state-of-the-art Resolution 1 Many scientists consider SDr commonly referred to as Star Wars incapable of achieving its stated goals and dangerously PERCEPTIONS OF destabilizing Participation by universities and GOVERNMENT FUNDING professional organizations lends a spurious FOR RESEARCH IN MATHEMATICS scientific legitimacy to it Therefore the AMS will lend no support to the Star Wars program Attitudes have changed for the worse in In particular no one acting as a representative the last two years At present academic of the AMS shall participate in efforts to obtain mathematicans are generally not very friendly funding for Star Wars research or to mediate to government or to DoD though many think between agencies granting Star Wars research that they should get support through the National Science Foundation almost as a matter money and those seeking to apply for it of right This was evident at the AMS business meeting during a debate on DoD Resolution 2 The AMS is concerned about funding of academic mathematics SDI and the the increasing militarization of support for proper relationship between DoD and academic mathematics research There is a tendency to mathematics Opponents of DoD pretty much distribute this support through narrowly focused controlled this meeting though in fact it was mission oriented programs which circumvent poorly attended and only a few people were normal peer review procedures This tendency unless checked may skew and ultimately injure involved in the debate mathematics in the United States Therefore Agency personnel interested in our recent those representing the AMS are requested to effort to build bridges to academia should study direct their efforts towards increasing the the debate to learn what does not work and fraction of non-military funding for what problems must be overcome mathematics research as well as towards increasing total research support On the positive side NSA attained Corporate Membership in AMS at the Council While the action parts of these resolutions meeting attended by Alberts Liebler and are mild and apparently designed to achieve Morris Selfridge spoke on our behalf during consensus the preceding rhetoric is not The the discussion on our membership citing the first two motions were to put these resolutions favorable impression he gained at the recent on the Agenda at the Salt Lake City Business NSA Conference He also responded to a query Meeting in August so that they may be about what NSA does drawing on information formally adopted The last two motions were to he had gleaned at the NSA Conference put the AMS on record as being in favor of the resolutions THE AMS BUSINESS MEETING AND THE DEBATE ON DoD FUNDING OF MATHEMATICS The board voted to call for a major increase in the NSF budget and for the AMS to sponsor mathematics competitions for students Under the rules of the AMS before a motion can be adopted it must be considered at two business meetings At the first it is voted onto the Agenda of the second and at the second it may then be formally adopted Accordingly there were four motions at this business meeting relative to the following two resolutions President Mostow decided to end the meeting on schedule so he placed severe limitations on time for debate Ulti mately all of the motions passed by slim margins Probably none of the motions would have passed had the rules been adhered to The few vote counts were of debatable accuracy Attendance at the business meeting was about 200 a tiny fraction of the people registered at the meeting not to mention of the AMS membership About 400 AMS members had signed petitions endorsing the resolutions Summary of points against the motion There should be multiple modes of funding DoD has the ability to take risks 1st Issue 1987 o CRYPTOLOG o page 22 POR OPPlelAI USB ONLY DOC I D 'B lrPIlQJ s increasing as fast as military funding o How can anyone object to doing research to see if SDI is valid Summary of points for the motion o Neither motion inhibits individuals from doing SDI work o DARPA money unbalances mathematics towards mission-oriented work and encourages students to enter some fields at the expense of others o Priorities should not be set by a military that lacks an overview of mathematics interdisciplinary research should not be allowed to interfere with traditional modalities According to him DoD program managers who understand science have little discretion or influence on scientific policy He sees that policy makers operate at a level where nonscientific issues are paramount and where the mathematics budget is negligible He opposes making such a structure responsible for the fate of 40% of federal support for mathematics He believes that when the nation's interests call for adjustments in the funding of mathematics the academic mathematicians can talk rationally to NSF But whom can we address at DoD who has the power knowledge and commitment to basic science to make appropriate changes o SDI thwarted disarmament at Reykjavik A second speaker believes that military research does not belong on university campuses for military organizations are o The dependence of computer scientists on hierarchical and secretive The size of military DARPA is excessive and harmful grants compared to other sources of funding distorts departments creating divisive minio Suppression of factorizations of integers shows empires influencing allocation of limited DoD's wickedness resources and setting priorities for graduate student manpower The research initiative During the debate in the evening people approach tends to give givers and takers a spoke both for and against the resolutions but common interest in disguising research failures predominantly in support of them There is more than enough military research One speaker focused on proper institutional being done already most of which does not policies in DoD funding He supported national improve military effectiveness security as a valid goal but he stated that He cited evidence that I loney from DoD military programs in support of research are would be used to quiet mathematicians from highly mission-oriented secretive and rely on criticizing DoD programs Military research scientific evaluation procedures of variable adds to a threatening atmosphere particularly reliability In his opinion AMS should do between the US and the USSR Military nothing to foster substantial funding of funding will change mathematical culture mathematical research in this military mode leading it towards classified research and ultimately weapons design Mathematics is a What may be more acceptable is defense poverty-stricken discipline and we face a programs in civilian mode operated with a shortage of mathematicians soon This does not reasonable variant of peer review and mean we should sell out If and when a unclassified in research areas selected for their consensus can be reached the Society should general relevance to the funding agency but lobby against military funding with little direction imposed on the individual investigators He spoke of the dangers of A third speaker gave a reasoned defense of political abuses i e non-scientific pressures accepting DoD help He pointed out that others on investigators or political exploitation of their wanted the government's money and that the participation without their consent as he said government need not beg mathematicians to has happened in SDl In the long term he accept financing He argued that priorities are wants a realignment of federal funding so that set by the civilians in Congress and by the the preponderance will come from NSF-like President and that if mathematicians want Agencies He believes that recent initiatives in more money they must show that they are valuable to society on the latter's terms If o Would SDI be wise even if it was possible 1st Issue 1987 o CRYPTOLOG o page 23 POR OPI I6It b VBS ONbY I DOCID 4010012 society valu es research of practical importance to the national security then mathematicians ought to consider this He said he was very disappointed with the business meeting the appearance of trying to set priorities for academic mathematics The MAA business meeting was extremely poorly attended In fact after the resolution was passed changing a quorum from 25 to 50 a jokester in the audience pointed out that the business meeting no longer had a quorum and so had put itself out of business Several proposed changes in the bylaws were routinely passed A fourth speaker made a passionate statement about the potential damage to the national defense if the resolutions were to pass Such resolutions may discourage mathematicians from dealing with DoD Another speaker said reliance on military money is a threat to mathematics He INFORMAL CONVERSATIONS P L 86-36 compared taking DoD money to taking money from the Nazis He later took this back He I had informal talks with Askey and Nevai mentioned the threat to cut off money to people on approximation of the tails of distributions critical of DoD programs Still another speaker referred to gold-plated toilet seats cheating Several of us L - _--- _ -- -_- -- __ and inequities in grants He said National and I - talked to a leader of the faction in favor Defense is a euphemism for Imperialism At of the anti-DoD resolutions I spoke to him best military spending is money down a rat about the Agency's new effort to strengthen hole academic mathematics and attempted to convince him that we are human A New Jersey woman said she had been told by a non-mathematician that she should Several times I talked to Persi Diaconis not teach mathematics because it will help blow who was in the Harvard Statistics Dept in up the world She asked can we change the 1972 while I was getting my degree in the perception of the world that mathematicians Mathematics Dept I asked him about some of help the military With this someone my moments problems Kailath remembered attempted to recruit more spokesmen for the me from a briefing he gave at Fort Meade DoD's side But since we saw which way the wind was blowing we felt it pointless to I Ispoke at length to A R respond and left the meeting Calderbank of AT T Bell Labs I supplied Calderbank with several references All this was rather discouraging to those in relevant to his work and informed him that favor of the new Agency effort to improve one of his results had been anticipated in a relations with academia The attitude of paper by At the academics at Louisville two years ago was more urging of Marshall Hall Calderbank is working in tune with the Agency's present policy of on the existence of a design with parameters collaborating for the common purpose of v k 28 10 5 improving the health of mathematics in the United States Those academics at this Don Rawlings of California Polytechni al meeting who made their views known were State University gave a talk on an extended generally of the opinion that the nation owes Simon Newcomb problem ltook them support that the support should be from him aside and educated hirn about a very NSF and that money from DoD is more likely powerful matrix technique published long ago to harm academia than help A substantial by Becker and of which Rawlings seemed minority still view DoD as the enemy of unaware humanity ltalked briefly to Ken Johnson of It seems clear however that DoD ought to Penn State who gave a very interesting tJ l lk give some thought to the effect its grants will on quasigroup characters relevant to nonat elian have in mathematics departments Fourier analysis Schur rings and associat ion Departmental politics may be as important a schemes In recent papers Johnson and Srilith consideration as national politics in making generalize much of character lbeoryr-- to _ _ these grants work DoD should try to avoid association schemes Johnson gavelL -_ I I I I I 1st Issue 1987 o CRYPTOLOG o page 24 Felt ep 'ICIAI USB eNLY P L 86-36 DOCID 4010012 several of these papers copies of which are available upon request Eileen Poiani of St Peter's College in New Jersey has worked at the national level on issues involving women in mathematics She was interested to learnfroml Ithat NSA has many highly qualified women among its professionalriiathematicians and asked if some of them would be willing to participate in her projects P L 86-36 At the NSA Conference Askey asked me to look up Nevai of Ohio State He was the organizer of a special session I asked Nevai about using orthogonal polynomials to approximate tails of distributions He asked me to send him my paper Askey and I talked several times At the NSA Conference he had argued for orthogonal polynomials as the solution to estimating extreme tails The threeterm recursion allows one to find orthogonal polynomials using the metric from the quadratic form dictated by the given moments In theory if the moments determine the distribution the set of orthogonal polynomials determines the distribution The practical utility for tail estimation is unclear I Lax had toldl at the NSA Conference that mathematicians have new and better ways of doing extreme tail approximation that statisticians have not yet learned Accordingly I cornered Lax and asked him what methods he had in mind Lax thinks he has a clever way of finding the roots of orthogonal polynomials without the expense of generating the polynomials See Diaconis' talk DR GRAHAM'S ADDRESS President's Science Advisor and Director of the Office of Scientific and Technical Policy OSTP W R Graham stressed the need for educational reform interdisciplinary cooperation and lobbying for more government funding The audience was tiny which mayor may not reflect the popularity of the US government at this conference It certainly reflected the availability of free beer at an Academic Press dinner held at the same timeD Graham spoke first on government investment in mathematics stating that the government is used to people lobbying for support and feels no need to force its money on people He stated that support to mathematics 1st Issue 1987 POR is based on the premise it is crucial to future economic and military security DoD is finally willing to give independent support for academic research a healthy sign Mathematics should involve itself in interdisciplinary efforts that prove its value In Graham's view the health of mathematics depends on finding ways to apply it to the benefit of the non-mathematical public He believes mathematics teaching in high school is poor We expect little of our schools and students so we cannot hope to compete with foreigners The Mathematics Science Education Board should help Demographics will be an increasing problem in getting enough good teachers and Agency employees Decentralization may make reform difficult RECRUITING The Agency recruiting effort was apparently quite successful About thirty people were interviewed This included most of the people who seemed to be good prospects Of course only time will tell which of these people will end up on board MATHEMATICAL HIGH POINTS A short course on Moments in Mathematics included among other things the following o Diaconis' work on Wyner's encryption method for speech over telephone lines and his ideas on attempts to extend Chebyshev's bounds o Kailath's talk on speech synthesis lattice filters fast methods of finding Cholesky and QR factorizations transmission line theory and a parallel algorithm for decoding certain codes Of the other lectures given the following were of particular interest o An interesting talk on radar by J Michael Baden Marvin N Cohen of Georgia Tech o McEliece on neural networks They can be used in theory to find local minima of NP complete problems and as storage devices o A talk by Durrett on modeling for a variety of growth and decay situations One wants parameter values for which a process dies out or lives CRYPTOLOG page OFI leIAL 25 USB ONb I DOCID 4010012 o Taiks by K S McCurley on sieving the positive integers by primes by Tapia on an unsuccessful attempt to use Karmarkar's methods to do quadratic programming and by R A Mollin on the class number of certain real quadratic fields SOME DETAILS Richard Durrett Cornell Univ Crab GnL S Measles and Gypsy Moths An IntNxt tction to Modem Probability A preprint of Durrett's paper is available The subject is five different models for interacting particle systems One asks in each case under what parameter settings the phenomenon survives forever or dies out 1 The Oriented Percolation Model 2 Richardson's Model Sites become occupied with a probability depending on the number of occupied neighbors 3 The Measles Contagion or Forest Fire Model similar to no 2 except there are three states tree burnt whole or on fire 4 The Gypsy Moth Spread Model similar to no 1 except that time is a continuum 5 The Crabgrass Spread Model on a lawn Z M d where M is a large integer J Michael Baden Marvin N Cohen Georgia Tech An Application of Undergraduate Mathematics Yielding Interesting Re Illts in Radar Research Cohen gave an introduction to radar with applications to ground mapping surveillance tracking and signature acquisition Long pulses give long range but weak resolution Pulse compression involves coding the long pulse to improve resolution Cohen explained filtering to handle sidelobes caused by correlations out of synch One wants to minimize both the integrated sidelobe level and the peak sidelobe There are 8 Barker codes that have optimal sidelobes But it is now known there are no Barker codes of length more than 13 and less than about 10000 One wants codes of length in the thousands One can combine Barker codes to obtain longer codes Also there are Golay pairs They are ideal except that movement of the target or any sort of instability kills them Baden uses a method involving mismatched filters to reduce the sidelobe The filter depends on the signal In practice one designs the filter with a sample of the signal and assumes this will be close to optimal One has to solve a linear system which by a physical argument is invertible Baden would like a mathematical reason why the system can be solved Mismatched filters do much better than the alternatives especially as the filter gets longer Charts allow one to select the proper filter length to get the desired sidelobe characteristics Marc Culler Univ Illinois Chicago Free groups Trees and their Automorphi3m This talk on geometric group theory included a reference to work of Ken Moss Given a free group F one tries to describe the outer automorphisms OUT F On the other hand following Thurston's work one seeks to characterize the points at infinity needed to compactify a topological space The connection between these two problems is a generalization of the concept of tree Robert J McEliece California Inst of Technology The Capacity of Neural Networks This is joint work with Posner Rodemich and Venkatesan Neural networks are arrays modeled on the brain Each neuron has a binary state firing or not firing It decides to change state on the basis of the sum of inputs from all the neurons to which it is connected There is a threshold function for computing the new state depending on a symmetric matrix T The initial state is called a prompt At any given time the node most urgently wanting to change is the one that changes with some tiebreaking procedure If the prompt is - 1 and we multiply this by matrix T and obtain for example the vector -1 5 -1 5 3 3 then nodes 1 2 3 4 want to change sign while nodes 5 and 6 do not The ones most urgently wanting to change are 2 and 4 so one of them is selected If T is symmetric with zeros on the diagonal this process will always converge to a fixed point a local minimum of energy In theory one can build a neural network to find local minima of NP-complete problems such as the Traveling Salesman Problem McEliece described how a network can be used as a storage device One designs a neural network with matrix T chosen so that the words to be remembered are fixed points However there may be other fixed points to complicate matters and other problems may arise Neural networks are easy to build and to update R A Tapia J E Dennis Rice A M Morshedi Shell Development Polynomial-Time 1st Issue 1987 o CRYPTOLOG o page 26 FeR OPI'IGI b lJiiE O LY DOCIDilrJ9 1Q Q l2 1 __ L _ -' ' mlllf1 based on Karmarkar's Linear Programming Approach Ye Tse and Kapoor Vaidya tried in 1986 to use Karmarkar's methods for QP Their efforts are unsatisfactory for different reasons The authors have tried again also with unsatisfactory results but for a bizarre reason R A Mollin Univ of Calgary Cla88 Mollin provides sufficient conditions for the non-triviality of the class numbers of certain real quadratic fields number of real quadratic fieldS K S McCurley USC D A Goldston San Jose St Sieving the Po8itive Integers by Large Primes If Q is a set of primes let 'I' x y Q be # n I n x and pin p EQ p y The authors prove an estimate for 'I' x y Q in terms of a generalized Dickman function extending work of Hildebrand This yields an asymptotic estimate for W x y Q One result is that removing small prime factors in sieving has less than the expected effect SHORT COURSE ON MOMENTS METHODS I attended a three-day short course on Moments in Mathematics In general the experts at the course tended to approach moments from a function-analytic point of view and had relatively little experience with statistics The course served as a general survey I particularly liked Kailath's applications-oriented talk Henry J Landau Bell Labs The Cla88ical Background Landau gave a survey of moments problems and applications beginning with the Hamburger Moment Problem find a measure on the real line with a given set of moments and determine if the measure is unique the Truncated Moment Problem only 2n moments given and the Trigonometric Moment Problem the equivalent existence problem but with complex moments on a circle - uniqueness following by Fourier transform theory A necessary condition for existence of a solution to the Hamburger Moment Problem is that a certain quadratic form be positive semi-definite In this case it defines a scalar product for polynomials and so a family of polynomials orthogonal with respect to this product He showed positive semi-definiteness is sufficient for existence Landau discussed generalizations and applications J H B Kemperman Rutgers Geometry of Given a set of measures we can map them to points in n-space by taking each measure to the point the moment point with coordinates the given n moments For measures on an arbitrary measurable space more general geometric constructs can be considered Given a finite number of moments one wishes to describe geometrically the convex set M of all measures having these moments or at least to give accurate bounds for the integrals of particular functions as the measure ranges over M Or one may require that some of the moments are not prescribed exactly but only fall within given bounds or on a line in n-space Or given a set A fix moments and ask for the largest or smallest mass for A consistent with the moments Such problems are conceptually simple but hard the Moment Problem Donald Sarason UC-Berkeley Moment Problems and Operators in Hilbert Space The connections between classical moments problems and the theory of operators in Hilbert space were recognized early on Sarason described some of these connections For example a onesided version of the trigonometric moment problem the so-called Nehari Moment Problem is related to the theory of Hankel operators A proof of the sufficiency of the positive semidefiniteness of the Hankel matrix for the Hamburger Problem follows from the Spectral Theorem for Self-Adjoint Operators One can give function-theoretic conditions for the solution to be unique Thomas Kailath Stanford Signal Proce88ing Applications of Some Moments Problems Trigonometric moment problems are connected with positive definite matrices orthogonal polynomials and classical function theory These subjects all are related in turn to a variety of signals processing problems linear prediction inverse scattering digital filtering etc Kailath illustrated the interplay between moments problems and signals theory Starting with linear predictive coding methods as applied in speech synthesis the way modern American toys have learned to speak he described the role of Szego polynomials in providing a nice hardware implementation The issue of parallel implementation of image processing brings in an algorithm of Schur This leads to a new class of digital filters called lattice filters that have excellent robustness to finite precision implementations 1st Issue 1987 CRYPTOLOG o page 27 FOR OFI ICIAh USB ONLY DOC I D n JoR s13 ast method for finding Cholesky and QR factorizations of Toeplitz matrices Other applications are to transmission line theory layer-peeling algori thms for inverse scattering and parallel algorithms for decoding Reed-Solomon and BCH codes Christian Berg Univ Copenhagen Denmark The Multidimensional Moment Problem and Semigroups The 2-dimensional moment problem consists in finding necessary and sufficient conditions for a sequence s m n m n 0 1 to be a moment sequence for a nonnegative measure on the real plane More generally in the n-dimensional case one would like to characterize moment sequences among all sequences characterize which are determinate and give a complete description of the set of measures with a given indeterminate moment sequence The theory of moments may be viewed as harmonic analysis on semigroups determination of the zeros of the associated orthogonal polynomials This is feasible for a small number of moments but is quite difficult in general Are these bounds on all distributions with a fixed n moments useful Probably the smoothness problems mentioned above occur generally Diaconis discussed Hausdorff's Moment Theorem which addresses the question when is a sequence the moment sequence for a distribution on the interval 0 1 and De Finetti Exchangeability Roughly De Finetti's theorem says that a distribution on sequence space is exchangeable if and only if it is a mlxture of Bernoullis in a generalized sense This theorem is equivalent to Hausdorff's Moment Theorem One can relate extensions of de Finetti's theorem to moment problems and so generate new problems MISCELLANEOUS Persi Diaconis Stanford Moment Problems in Probability and Stati3tics Statisticians use moment estimators as basic in applied work For example Diaconis said that the distribution of traces on the n-dimensional orthogonal group has the same first 2n 1 momen ts as the standard normal distribution In the application which was a proposal by Aaron Wyner of Bell Labs to encrypt speech for phone lines using random orthogonal matrices this was bad news Diaconis and his co-workers were trying to approximate Haar measure for the orthogonal group using iteration of random reflections A reflection fixes a hyperplane so log n reflections will fix too much and the trace of the iterated operator will be too large One needs n log n reflections to get the trace right Many exhibitors were present and distributing literature Literature and demonstrations on word processing were particularly evident Mike Spivak showed his own commitment to TEX by typing people's equations in great numbers Among conferences advertised was the First Canadian Number Theory Society Conference at Banff April 17-30 1988 A tentative list of speakers includes Erdos Guy H Lenstra Pomerance Ribenboim Selfridge Washington H Williams and Zassenhaus The AMS awarded Steele Prizes to Donald Knuth for his book Rudolf Kalman for papers on filtering and linear dynamical systems and Saunders MacLane for cumulative Diaconis mentioned Chebyshev's bounds on influence especially on homological and a distribution with the same first n moments as categorical algebra The Cole Prize in Number the standard normal distribution There is a Theory was shared by Benedict Gross Dorian bound on the error but it is attained by Goldfeld and Don Zagier for work on the class discrete distributions If one looks at smooth number problem for imaginary quadratic fields distributions the bound is very poor Diaconis As for MAA awards the Chauvenet Prize went would like more work on what can be said with to J H Wilkinson the Allendoerfer Awards to smoothness assumptions I mentioned that Bart Braden and Sol Stahl the Ford Awards to robustness considerations both on the moments Jeffrey C Lagarias and M E Taylor and the being equal and on smoothness must be Polya Award to P J Davis 0 studied What if one drops normal Diaconis and I have both been trying to get people to work on such problems Editor's Note Wejust learned that the resolutions pas ed during the AMS business meeting were Diaconis said that the modern approach to subsequently ruled out oforder by the General bounding variability among distributions with a Council It is expected that there will be a prescribed n moments leads to numerical referendu m on the motions after January 1988 1st ssue 1987 CRYPTOLOG page 28 POR OfT CIAI USB ONLY DOCID r40rOOIZ UNSA PUZZLE P L b 86-36 '_1 T12 The occurrence of NSA as a letter sequence in English words is rather uncommon Some of the rare examples are shown below Fill in the blanks to reconstruct the NSA words defined A Crazy _n s a B Land of Opportunity _ _ _ _ nsa C Lacking sodium chloride _n s a D Pay for damages _____ n s a E Miniature tree __ n s a F Unable to be satisfied _n s a G Across the ocean ___ n s a ' H Middle Eastern inn n s a __ I Unhealthful as a climate _n s a J Not spoken _n s a K Search thoroughly __ n s a L Deny or contradict ___ nsa M Not paid regularly _n s a N Relationship by descent from a common ancestor __ n s a O Unclean _n s a P Result of a bank visit ---n s a Q Disagreeable or disgusting _n s a R Type of roof __ n s a S Sail on a square-rigged vessel ___ n s a --------- T Able to combine with other _n s a substances 1st Issue 1987 CRYPTOLOG page 29 'on O TleIA use aNI ' I I DOCID 4010012 IIAPdDlE VIA COMINT C1IANNElS ONLY NOT RELEASABLE TO CONTRACTORS -5EEREf--- 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