NSA on 10-17-2012 ursuantto E O 13526 MDR Case # 5477 l3 lDrnVU UJUil31I1 Ull DlEUll UlQ Wl UJlDl J' WUJlE I1UllIJlD lIlUJUVUllIlUJl1 r lE 7 3rd Issue 1988 ip L CHANGES IN AGENCY REPORTING SIGINT REPORTING PRINCIPLES UPDATE ON PACKET RADIO o o EMBARGOED INFORMATION o o o BULLETIN BOARD o o o o o o THE SHORT BUT HAPPY LIFE OF - - 1 o o o o WIDEBAND TillS BO JI JMBNT JONTAINS eOBEWORB MA ERIAL EO 1 4 c LPOLO 86-36 1 1 r N C Gerson LANGUAGE BRIEF XHOSA o o o EXPO 88 o o o o o o o o o o CONFERENCE REPORTS MATHEMATICS o LINGUISTICS o COMMUNICATIONS ON THE LIGHTER SIDE o CALL FOR SPEAKERS o LEFTOVERS o o o o o o o o 'r I I I I 1 1 I r 86-36 1 3 3 4 5 6 9 14 16 17 19 22 27 28 29 CLASSIFIED BY NSA CSSM 123 2 DECLASSIFY ON OrigiAatiAg AgeAey' O@t@rminatioil Reqail ed NOT RELEASABLE TO CONTRACTORS I DocrD 4010015 - isliedbY Pl Te hniques and Standards igk J 3 PuBh$t R 3rd Issue 1988 I -- ItM 'i F DrrDRS COllection o o Computer Systl l 1 C' -- p ' d'Te i I ' ' C Spe ial Research Traffic Analysis 963-5877 963-1103 J i 963-3456 1 '- IooH 3 E t ' ' ' ' Information SCien e Information Security ' ' Intelligence Research Language ' ' Mathematics -' YELLOW PAGES FOR NSA 972-2122 963-3845 963-3057 f 963-5566 f What NSA needs is Yellow Pages just like the phone book showing functional responsibilities Or where to get information Or even how to found out how to get information What brought this to mind is a kind of inquiry that CRYPTOLOG often gets Here's a typical exchange Reader I'm looking for articles on Project XYZ It seems to me I read about it in CRYPTOLOG Editor Let me check Pause No sorry you saw it in The Cryptologic Quarterly Probably The two publications ore often confused Their number is 972- 2355 Vera R Filby 968-8014 A week Rpbert J Hanyok 963-435 I asses Reader About that Project XYZ CQ tells me that they didn't publish it either Do you latow where else I can try To submit articles or letters by mail send to Editor CRYPTOlOG PI HQ SAI87 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 cryptlg@bar1c05 bar-one-c-zero-five note no '0' Always include your full name organization and secure phone also building and room numbers For Change of Address mail name and old and new organizations to Editor CRYPTOLOG PI HQS SA187 Please do not phone Editor Try STINFO T513 Their number is 9688611 They can do a keyword search in classified documents and in open 3OU1'ces as well Reader Thanks By the way Someone from another agency is coming in for a briefing on the project I'm s tpposed to make the arrangements Where can I find out what I have to do It would be nice to say Look it up in the Yellow Pages which of course is rinted on yellow paper-or better yet in this day of automation is kept on line It could be limited to a few outlets one terminal per office something like the personnel files on the M204 for example Why not Contents of CRYPTOLOG should not be reprodu ed or disseminated outside the National Security Agency without the permission of the Publisher Inquiries regarding reproduction and dissemination should be directed to the Editor All opinions expressed in CRYPTOlOG are those of the authors They do not represent the official views of the National Security Agency Central Security Service FOR OFFICIAL USE ONLY DOCID 4010015 eONF'IB'8N'f'IAL CHANGES IN AGENCY REPORTING I------ J p L e-ee6 In November 1982 NSA with G Group at the lead launched a major initiative to move away from the traditional full-text translation format I- ----J to a journalistic format e-ee6 The basis for this change was multifaceted One was security related We had lived through instances of seeing verbatim quotes of some of our product in the media and needless to say even in an English translation form a quote in the media of a verbatim translation of target communications makes it much easier for a target counterintelligence effort to track our source and to take countermeasures While the journalistic format does not provide full-proof protection in the case of leaks or other compromises of our product it does give at least some additional security 86- 3 6 learns what the report covers and can decide whether to read the entire report In contrast when the information is published as a full-text translation the reader often has to wade laboriously through long-winded salutations and extraneous material to extract the important parts of the report 0 GGO Finally another reason 'for the switch to journalistic style was that we have only a limited number of foreign-language translators to cope with an ever increasin volume of interce t U Another reason for the move to the journalistic style was to provide a product Some L -_- ----_ _ better suited to the reader especially the reporters complained that the switch to the executive reader who has to scan rapidly much journalistic style detracted from their verbatim incoming material to determine quickly the key translation skills upon which they were tested elements of the reports The journalistic format for certification Some senior che ckers encourages the reporter to summarize briefly complained that the switch to the journalistic the hig ghts of the report at the beginning of format actually took more time because they the report Thus by reading the title and the could not feel confident of the reports done by introductory summary the reader quickly the junior linguists unless they first saw a full- -- _ --- - - EO 1 4 c P L 86-36 CRYPTOLOG page 1 eONF'IB'8N'f'IAL lIlA GOMUT' ' GIIA1'B'lB bS m bY 3rd Issue 1988 W Mln 8 P L 86-36 P Q 1 4 c d P L 86-36 DOCID 4010015 01 4 G9NFIQ N IhL text translation upon which the report was based U But gradually those growing pains have been overcome Our reporters and checkers have adjusted to the journalistic style and now most feel quite at home with it preferring it over the old style because it makes themthink more about the story they want to convey and how they want to convey it CUSTOMERS' REACTIONS C CCO As the new Agency initiative began to develop momentum we got mixed reactions from readers Some were uite leased with this new format WHERE WE ARE HEADED EO CeO Any changes as dramatic as the Agency's move to the journalistic style for our lproduct would be expected to take tIme as we progress toward our objectives New reporting policies and procedures have evolved as we made adjustments to come up with a more efficient production process and to provide the best product possible to our customers Sometimes communication with our readership with our Second Party agencies and even within our own ranks has not been as good as it should have been But with the patience and persistence of top Agency officials managers at all levels and reporters at the Agency and in the field we have made tremendous progress since setting out on this road over five years ago I L- ---l As time went on we encouraged customers to develop a direct relationship with our reporting elements to help answer any questions arising with our product POUO Our main effort now is to consolidate COORDINATION WITH SECOND PARTIES these changes into a new draft of USSID 300 in a way that will clearly document these and other changes for the workforce And we have established a DDOIntelligence Staff Product Improvement Working Group to provide a forum for within-DDO communications of the success and problems associated with this initiative and to surface other ideas on ways to publish better product with our limited resources We encourage the active participation of everyone concerned in these endeavors 0 ------------------- EO 1 4 c EO 1 4 d P L 86-36 EO 1 4 c P L 86-36 3rd Issue 1988 o CRYPTOLOG o page 2 eONFIDI 3N'JPYIAL HAUBI il3 VIA COMRff CIIMUfl3LS OHV DOCID 4010015 36 SEeRfi UPDATE ON PACKET RADIOS L U As a follow-up to my article Is Packet Radio Alive and Well on HF Cryptolog 3rd Issue 1987 here is some recent information U The following countries should be added to figure 2 countries whose amateur radio operators are active in amateur packet o Bahrain Bermuda Bolivia Bulgaria Egypt French Polynesia Gabon Greece Honduras Hong Kong Iceland India Ireland Israel Kenya Lebanon Lesotho Malaysia Malta Marinique Monaco Namibia Nepal New Caledonia New Guinea Nigeria Oman Pakistan People's Republic of China Peru Poland Qatar Senegal South Korea Soviet Union Surinam Thailand United Arab Emirates W stern Samoa Yugoslavia Zaire UNCLASSIFIED EO 1 4 c P L 86-36 3rd Issue 1988 o CRYPTOLOG o page 3 SfiORfitf SPOIiB 86-36 DOCID 4010015 EO 1 4 c P L 86-36 SHORH Release of Embargoed Information 86-36 P05 U This is a discussion of pertinent facts and regulations concerning the release of embargoed items to non-NSA contractors Keep in mind that the distinction between embargoed and unembargoed information applies only to information released to the non-NSA contractor to-OOO The key to identifying end-product that DIRNSA considers embargoed is that regardless of its COMINT category it bears caveats handling restrictions or channels markings other than or in addition to COMINT channels U DIRNSA's policy is to not release any embargoed material to a non-NSA contractor on a blanket basis In other words the sponsor must identify each and every specific document requested together with supportive justification A general request such as All information pertaining to technical intelligence on the Isystem is unacceptable I U Embargoed information falls into the following categories 0-000 End-product regardless of its COMINT category bearing any restrictive caveat such as NOFORN ORCON GAMMA Controlled Item PROPIN NO CONTRACT etc is considered embargoed information These are examples of restrictive handling and distribution exchange designators and do not purport to represent all possible versions of such restrictions U Any end-product protected in a special channel such as LOMA and or 3rd Issue 1988 P L 86-36 Ii L n J E Vl A TALENT-KEYHOLE instead of or in addition to COMINT channels U All technical data is likewise embargoed regardless of classification and or channels protection Technical data includes working aids technical support letters reports and manuals information of SIGINT production processes raw SIGINT data tapes of specific signals and tapes of electromagnetic spectra U Foreign encrypted signals are not repeat not releasable to non-NSA contractors without the specific approval of the Director or Deputy Director NSA To Request Release of EMBARGOED Information FOUO A request for the release of embargoed information to a non-NSA contractor involves the following steps POUO 1 The sponsoring agency's contract monitor must submit a written request electrical messagelletter memo to DIRNSA ATrN POSISAO Military sponsors should forward their requests to the cognizant intelligence command for review and validation with an information copy to DIRNSA POS SAO Upon POS SAO's receipt of the cognizant intelligence command's validation it will initiate processing action CRYPrOLOG page SECftET OOMHT OIIAn TELS 4 mTLY I I DOCID 4010015 Sli3eRIi3 F'OUO 2 The request following contain the a The title contract number and expiration date of the contract BULLETIN BOARD b The contract statement of work or information sufficient for NSAlCSS understanding of the contract's purpose WORD PATTERN LISTS c Certification of the contractor's need for access to the SIGINT infomation d Certification of the contractor's clearances e Security certification of the contractor's storage facility or the facility in which the SIGINT information will be processed IBM TO XEROX CONVERSION f Name and phone number of the contract monitor U 3 Upon receipt of the above P05 SAO will review the request and determine the appropriate action office generally dictated by what organization originated the material and or the OPI of the subject matter involved In most cases the OPI action office is the final authority regarding the releasability of the material requested and has the prerogative of mandating specific provisions or precautions to be adhered to in connection with the release The OPI is also responsible for advising P05 SAO concerning the decision to release of the requested information U 4 P05 SAO will then forward the Agency's decision to the requester in writing based upon the recommendation of the OP F'OUO NSA-originated SIGINT or SIGINT- related material may not be released to a nonNSA contractor unless the above process is observed and P05 SAO DDO's designated Central Point of Contact regarding release of material to contractors is aware of the request action U For clarification as to whether material requested for a non-NSA contractor falls into the embargoed or unembargoed category please call me on 963-54635 0 3rd Issue 1988 HA TBY l F'OUO A program to generate word pattern lists is available in G244 It produces patterns in the familiar form ABCDAC et al when run on an edited plaintext data base For information call Fred Pollnitz or Bob Silva 963-5380s U There is now a program to strip superfluous function codes carriage return and line feed from IBM floppies before conversion to the XEROX STAR format Preliminary tests show that it also works for IBM to Macintosh conversion This is a quick routine a matter of seconds that eliminates the customary time-consuming multi-step find-andreplace routines For a copy of the program and instructions call or writel P16 HQS 963-1103 P L 86-36 1 SOFTWARE VERIFICATION U The National Bureau of Stapdards has adopted a new standard for improving the accuracy and quality of software during development and maintenance Called Software Verification and Validation Plans ANSDIlIEEE 1012 1987 it was developed by IEEE We expect to see it published as a Federal Information Processing Standard in the near future For further information call or write I IP13D FAl' X II 96 8161 P L 86-36 WANTED CONVERSION FROM APOLLO U CRYPrOLOG is seeking to convert files prepared on APOLLO particularly those done in Tex to IBM DOS or Macintosh or Xerox STAR If you know how to do it please get in touch with the Editor l Ip1 HQS 963-1103 CRYPTOLOG page 8 6RIi3 P L 5 VEA OOMIU'i OIIAf'HiElLS ONLY 86-36 SEeKE'f EO 1 4 c P L 86-36 THE SHORT BUT HAPPY LIFE OF WIDEBAND ------------------------------ ----- ------ ----------- o z ' t iJ ---- - ' ' ----- -- -' There was a time when we had collection resources all around the world The - - - - -- ---- --- '-1- - - - - -- 043186-36 the duty less pleasant would make it more difficult to find people to do the job The only reasonable approach to this problem was tp develop some meu od of limiting the number of people to spend the money manning in those stations however was to become a source of concern for Agency management Collection activity as it was performed in those bygone days was manintensive and quite costly U In the early sixties another problem arose that exacerbated the concerns about station manning The French for whatever reasons they had began a run on the gold reserves of the United States by cashing in dollars Other nations not wanting to be left holding the bag when the U S ran out of money followed suit Several programs were started on a national level to halt the erosion and to bolster our country's reserves but they are not involved in this story U NSA as its contribution soughtmeans to limit its contribution to the gold flow One of the most significant factors identified was the pay that our field personnel spent ofT base That money went into the foreign economy as dollars that were at that time redeemable in gold U This was the issue How do we limit the spending of our personnel on the foreign economy Cutting their pay limiting ofT-duty time 01' restricting ofT-base activity were all equally unattractive as solutions We have llever enjoyed an overabundance of the skilled intelligent personnel required to man a SIGINT field station and any solution that would make 3rd Issue 1988 o CRYPTOLOG o page 6 SEl6REl P L 86-36 _ P L '-- - -- -- - ----- - -- -- - - - ---- - - - - EO 1 4 c IIA WU3 VIA OOloiH R OIIA HiELS ONp ' L 8 6- 3 6 DOCID 4010015 1 - 1 4 c L 86-36 SEOREl'f' ------------------1 EO 1 4 c P L 86-36 U ECSR has been considered by some people as an evolutionary stage in the development of remote intercept technology and that may be accurate But just as the tyrannosaur the 3rd Issue 1988 o CRYPTOLOG o page 7 SHORE'I' IIAP'fBbE VfA OOM lP'f'f' OIIAUUHLS m LY I I EO 1 4 c DOCID 401001 -36 diplodocus and the triceratops may be gone the gekko the alligator and the iguana are their descendants and still fill ecological niches So too does the ECSR system ot'today serve its collection functions - - - - - - - - - - - - _--- realized from the slower tape speed and the reduced footage that would be recorded and played There are no definitive answers to that yet U One additional advantage to the half speed modification would be the fact that the new version could probably be phased in more easily than could the introduction of entirely new technology Thus the cost of the option could be spread across a few years resulting in smaller impacts on annual budgets and less disruption to ongoing activity New technology would require a turnkey operation and while it - - - - - - - - - - - - - - - - - - could be phased in by theater it would U A new chapter is being written in this probably have to be done theater-wide in each story Just as the technique is more than phase twenty years old the equipment now being used is ten to fifteen years old and is beginning U There is one other disadvantage to the to show its age half speed proposal however It involves replacing the heads capstans and other parts U There are people in NSA who are on the recorders already in the inventory beginning to be concerned about the equipment With this you have new technology on what and the methods now in use and who are are still old machines We would be putting persuading themselves that new technology hundred dollar saddles on ten dollar horses should be developed to supplant the banks of The motors power supplies and other parts of recorders that are in use to record a little more the machines will just continue to get older than a Mhz apiece There is discussion of Worse yet when the parts that never break digital recording techniques techniques using start to break they will be very difficult and VCR technology and of techniques that retain expensive to replace-all out of proportion to the basic technology in use at present but their original value-if past experience with modified to record at half the speed used now other kinds of hardware holds true thus saving - in theory at least - head wear tape wear and tape volume Reduction of the EO 660 There is no good answer to what we tape volume in turn would also reduce tape should do about our wideband'resources storage space requirements shipping costs and Perhaps before we consider the answers to the expenses involved in rehabilitating and those questions we should examine other inspecting the tapes questions For example is wideband recording a'n for itself in intelli ence ains U Each of these options has one thing in common with the others They are all very expensive The least expensive of the options the half speed modifications has been tested - -- -_- -- ----' Is and seems to work well judging by the limited there any other way to derive what we n ow get experience we have had with it The cost for from wideband recording modifying each recorder would be in the neighborhood of $50 000 per unit and the U This has not been a research paper it modification also requires that new high has been some random ideas that have crossed the writer's mind The historical chronology performance tapes be introduced into the system Tapes now in the inventory will not may be fuzzy the costs cited may be serve inaccurate but the basic facts are facts and we are coming to a time when a decision should he U It has been postulated however that the made high performance tapes may be more abrasive Q- 1 4 c than the ta es used at present which would of U A ny suggest IOns p LOT o 8 6 - 3 6 course increase the wear on the recording and playback heads Would that offset the savings 3rd Issue 1988 o CRYPTOLOG o page 8 SEeRE HAnDLE VIA COMHof't' CIIAH'lof'ELS emLY DOCID 4010015 I SKGRB EO 1 4 c P L 86-36 N C G E R S o N R 6 'fiBS A-R'flCLEl IS CUd3SlF'f8B SHCRH'i' IIl CCO Hi JPYfS HN 'fY 3rd Issue 1988 o CRYPTOLOG o page 9 SIWRH WA lQbK V A GOMlP'l'f OHANP'lBbS mlb Y oP L 86-36 EO 1 4 c EO 1 4 d DOCID 4010015 P L 86-36 EO 1 4 c EO 1 4 d SBCRBqJ 3rd Issue 1988 o CRYPTOLOG o page 10 SBCRSqJ HAfiBLB ViA COMiU'f CHANNELS miLY I DOCID 4010015 P L 86-36 EO 1 4 c EO 1 4 d I 8I'lORH't' 3rd Issue 1988 o CRYPTOLOG o page 11 SHeKt 'f HAUDLE VfA COMEU't' CIIAUrq'HLS or LY DOCID L 4010015 86-36 EO 1 4 c EO 1 4 d 3rd Issue 1988 o CRYPTOLOG o page 12 SECRIi3'J' IIAPfBI J1i3 JIM COMnf'J' CIh r rE bS O r bJPY DOCID P L 86-36 EO 1 4 c EO 1 4 d 4010015 SIWRB'f 3rd Issue 1988 o CRYPTOLOG o page 13 SB6R8T HMIBI 8 VIA 60ltH f' F OII t H'fBLS DULY DOCID 4010015 L 86-36 TQP SBORH'f UMBItA U Xhosa is a Bantu language spoken by more than 6 million people in the Transkei and Ciskei homelands of South Africa which border Lesotho and face the Indian Ocean The Bantu family of languages numbers over 300 and is spoken in an area encompassing most of the southern part of Africa extending south of Cameroon and Kenya Xhosa was the first South African Bantu language to appear in print transcribed by missionaries in 1821 Xhosa is a tonal language with high low and falling tones Because the Xhosa people interacted with the Hottentots early in their history their language shares some of the peculiar clicks characteristic of the language spoken by the Hottentots people of a South African race related to the Bushmen Sotho also shares some of these characteristics although it has only a single click The clicks are represented by th e Iette rs X Q an d C and in combination with vowels and consonants form 20 click combinations Changing one click combination for another can change the entire meaning of a word LANGUAGE BRIEF XHOSA U U The X click or lateral sounds like an equestrian urging his horse on the Q click or palatal sounds very much like a cork being released from a bottle The e click or dental sounds like the nonverbal sound a person makes to express disgust U Xhosa is based on a system of concords or prefixes Depending on which of seventeen classes the noun of the sentence fits into the prefix of the noun will influence the form of the verbs adjectives and other parts of speech which are brought into relationship with the noun Examples are shown below Because of these concords Xhosa is extremely alliterative which makes it fairly easy to recognize in written and spoken form predominance of detachable particles which can be added to the word to vary its meaning It is a language filled with interjections and has ideophones as well Ideophones are descriptive and exclamatory utterances designed to complement the thought being expressed and are often imitative of the action described or of a sound associated with it U Although Xhosa is one of the principal languages of South Africa along with Zulu Tswana and Sotho it is not considered an official language in that white-ruled country Only Afrikaans-a bastardized version of 16th century Dutch-and English are officially recognized U In addition Xhosa is an agglutinative language as opposed to inflectional having a Unclassified EXAMPLES OF CONCORD All people want peace onke aQantu Qafuna uxolo The whole country wants peace onke ilizwe lifuna uxolo The whole nation wants peace onke i izwe ifuna uxolo All nations want peace onke i izwe ifuna uxolo 3rd Issue 1988 CRYPTOLOG page OP S130RH 14 UMBRA EO 1 4 c P L 86-36 I DOCID L 4010015 OP SHCRE'f' UMBRA Unclassified 1 4 c L 86-36 3rd Issue 1988 o CRYPTOLOG o page 15 'flOP SECRE'fl l JMBRA I DOClD 4010015 r -I I -I -I -I -I -I -I -I -I -I -I -I L ------------------- EXPO 88 19 October 1988 Friedman Auditorium See it now -I -I -I -I -I o I -I High tech for linguists o Custom-designed workstations for linguists I Online dictionaries Electronic gazeteers High-resolution displays Neat inventions Useful gadgets - I -I -I o I o Sponsored by The Office ofthe ChiefScientist and the Natural Language Working Group a consortium oflinguists computer scientists and engineers from A B G T R and Z I ------------------- J 3rd Issue 1988 o CRYPTOLOG o page 16 FOR OFFICIAL USB ONLY I DOCID 4010015 attendance is indispensable More than one representative should be sent when possible CONFERENCE REPORTS This may be an appropriate place to give a view on the value to NSA of attendance at meetings of the present type I believe that the dominant factor is the educational and motivational value to the individual attendees They are exposed to a lot of new information by enthusiastic researchers from the US and the world they make contacts and are made to feel a part of a larger mathematical community This will be reflected in better job performance Attendance at meetings is an appropriate and economical mode of training for those of us who are already steeped in formal degree-oriented schooling The individual attendees will be better trained better motivated and have improved morale Benefits like this have the biggest potential payoff THE TALKS Nineteenth Southeastern Conference on Combinatorics Graph Theory and Computing 15-19 Feb 1988 Louisiana State University Baton Rouge Reported by I P L 512 86-36 I attended this conference to present my paper A New Combination Generation Method to keep abreast of developments in a vital area of mathematics and to support Agency reeruitment of professional mathematicians I There were five invited presentations and about 180 fifteen minute talks in three parallel sessions As conference proceedings are scheduled to appear in Congressus Numerantium published by Utilitas and also because no earth-shattering new results were announced I am limiting myself to comparatively few remarks about the talks themselves Each of the five days was highlighted by an invited instructional lecture of one or two hours All of the invited addresses with the exception of Frankl placed considerable emphasis on whether the problems addressed are NP complete or had efficient algorithms I interpret this to be an indicator of current fashion They were Cycles in Graphs and Directed Graphs by Carsten Thomassen of the Technical University of Denmark Fast Algorithms for Convex Polygon Problems by Maria Klawe of IBM Almaden Research Center Current Developments in Cryptanalysis by Ernest Brickell of Bellcore and Sandia Labs Old and New Problems in Finite Sets by Peter Frankl of Bell Labs and the University of Paris and Recent Developments in Ramsey Theory by Vojtech ROdI of Bell Labs and Czech Tech The last two speakers are mentioned as possible successors to Erdos Erdos himself was to have talked but skipped off to Australia at the last moment so Frankl spoke instead All of CONCLUSIONS AND RECOMMENDATIONS these talks were of high quality The Southeastern Conference addresses a type Klawe studied a scheme whereby certain of mathematics of very special importance at inequalities related to convexity propagate NSA namely combinatorics Hence NSA 3rd Issue 1988 o CRYPTOLOG o page 17 FOR OFFI6IAL USE ONLY I DOCID 4010015 through a matrix Totally monotone is her term for matrices in which this happens This structure was exploited to construct certain efficient algorithms Her scheme appeared similar to one which I had employed in computing what I call subsequence numbers I pointed this out to her and gave her a reprint of my paper Discrete Mathematics 16 1976 123-140 Brickell's talk surveyed a lot of open-literature cryptanalysis very fast I am told the presentation was not much different from previous talks he has given Nevertheless since there is probably interest in what he had to say I have obtained copies of his multitudinous slides which he was kind enough to send me He kept repeating that much of this material would be included in a paper to appear in the Proceedings of the IEEE written with Odlyzko Frankl's talk was fairly short and included problems concerned with the existence and construction of critical e g maximal or minimal families of subsets of a finite set with various stated requirements on how the subsets in the family had to be related one to another Sperner's theorem for example concerns the case where the requirement is pairwise incomparability In most of the cases Frankl discussed however the requirement concerned intersecting or cross-intersecting between sets in the family where to intersect means to overlap As an illustration he gave a result which says if a collection of subsets of an nelement set has the property that any two subsets in the collection intersect then that collection cannot have more than 2 1 subsets in it It is easy to verify that the stated bound is attained sums of characters I have a copy of his preprint Chung's principal result translates Katz's lemma into a bound on the diameter of a k-regular graph in terms of the eigenvalues of its adjacency matrix I have the abstract of her relevant papers The papers Chung sent me are as follows a an' estimate for Character Sums by Katz 5pp used in b b Diameters and Eigenvalues 19 pp c The Average Distance and Independence Number 12 pp d A Dynamic Location Problem for Graphs with Graham Micheal E Saks 36pp e Diameters of Graphs Old Problems and New Results 26 pp 0 Quasi-random Graphs with Graham R M Wilson 35 pp These papers are undated I will try to supply copies to anyone who expresses interest Chung's paper ec verifies a conjecture of GRAFFITTI GRAFFITTI is a computer program and the conjecture is that in any connected graph the independence number is at least as large as the average distance between vertices GRAFFITTI was wrjtten by another attendee at the conference Siemion Fajtlowicz Math Department University of Houston GRAFFITTI has apparently enabled Fajtlowicz to look at a large number of graphs and to formulate conjectures on the basis of the results He has sent me a recent update on his conjectures this has been going on for several years which I will supply on request Harary mentioned Fan Chung in his 15minute talk on sum and difference graphs as having character theoretical techniques for Fajtlowicz may be a valuable source for obtaining bounds on the diameters of graphs graph algorithms even in cases where efficient Since the latter was in attendance I sought her no really efficient algorithm is possible ie in out for details She is married to Ron Graham cases where the problem is NP complete This of Bell Labs It quickly became apparent that is because GRAFFITTI has had to contend with her work is pertinent to routing problems in such computations for example in the case of distributed memory processing arra a bein the determination of a largest clique in a given considered for HORIZON at SRC r- --- -- -- ----_---Jgraph FajUowicz told me he would be happy 1 -_---' She sent me quite a collection of preto supply me with this particular algorithm prints including a crucial lemma by Nicholas I didn't actually ask for it however M Katz which appears to be significant in its My own talk on Mardi Gras morning went own right Katz obtains a bound on certain very well It was a variant of my talk at last P L 86-36 3rd Issue 1988 CRYPTOLOG page 18 F6ft 6FFfefAL tfSE 6NLY P L DOCID 4010015 year's NSA Mathematical Sciences Meetings so doesn't need elaboration here I did meet a fellow interested in this sort of thing who seems to have done new and interesting work in this area He is Frank Ruskey of the University of Victoria in British Columbia He supplied me with a number of recent reprints which I am en thusiastic about but which I haven't yet had a chance to assess in detail RECRUITING people in the town of Baton Rouge in the shops and parks and streets One towns-family even invited me to ride with them to New Orleans for Mardi Gras but I decided to go with the conference crowd I would enjoy going back 0 The Linguistic Society of America San Francisco December 27-29 1987 Reported In accordance with policy I identified myself as an NSA employee although my name tag read Department of Defense I got permission from the conference organizing committee to leave recruitment material on the table reserved for new publications I left copies of regular NSA technical recruitment brochure and of Brent Morris's brochure I also placed a couple copies of the Proceedings of the NSA Mathematical Sciences Meetings One copy of each I marked to remain on the table throughout the conference The Proceedings attracted a lot of interest Most of the time someone had it open and was reading it ath Panel I suggestedl Executive as an NSA contact for math employment inquiries I stapled one of his business c ards inside the table copy of the Tech Brochur and the Meetings Proceedings I think these measures were both conservative and effective AFTER HOURS Having completed my own presentation on Tuesday morning I was in an ideal frame of mind for the afternoon and evening trip to New Orleans for the famous Mardi Gras celebration That day the conference adjourned at noon Having been prepared by warnings about staying out of dark alleys and not keeping a wallet in a back pocket Bert Hartnell Saint Mary's University Halifax and I spent most of our time going up and down Bourbon Street Bert and I managed to wend our way intact back to our bus at 11 PM I was worried that I had led Bert to ruin but his talk the next morning was great I was treated very well by local people There may actually be something to the notion of southern hospitality This does not refer only to people connected with the conference who sort of had to be polite but especially to the b y 1 1 P16 The annual meetings of this society are the principal means of reporting developments in linguistic theory and applications Keeping abreast of developments is important to NSA because of the increasing application of linguistics to Agency work I hope that other NSA people with training in linguistics will be able to attend future meetings As proceedings are not published I am summarizing the talks I attended arranged by topic SYNTAX Ronald Kaplan and Annie Zaenen Xerox Parc Crossing Dependencies in Germanic Languages and Functional Precedence This paper gave an LFG account of infinitival and participial constructions in West Germanic languages The LFG account allowed them to posit that arguments of an infinitival complement follow those of higher verbs without having to introduce them in the same phrase-structure rules P L 86-36 o Nigel Fabb University of Strathclyde NonRestrictive Relative Clauses and D-Structure Adjunction Scope tests show that a nonrestrictive relative clause is not c-commanded by its antecedent while a restrictive relative clause is However tests involving binding show that the antecedent does not c-command parts of the non-restrictive clause Fabb examines two different options for explaining these data DISCOURSE Joan Levinson City University of New York The Linguistic Status of the Orthographic Text Sentence CRYPTOLOG page 19 POR OPPIEURlAb USE Qf bY 3rd Issue 1988 86-36 I I DOCID 4010015 This paper discussed factors which affect the distribution of marks of sentence punctuation Levinson first showed that neither the boundaries of sentences nor the sentence internal punctuation is syntactically specifiable and then offered a theory of information grouping to account for punctuation This theory suggest that punctuation is a guide for how to read a text and may reflect some kind of intonational grouping In the sentence John said that Mary is pregnant a present tense in a verb complement clause is embedded under a past tense thus violating the sequence-of-tense rule and causing the time of the embedded clause to contain both the time of the matrix and the speech time Ogihara showed the deficiences of a syntactic solution Enc forthcoming and offered a semantic solution Jack Dubois UC Santa Barbara Discourse Promotion and the Two-Track Hypothesis PRAGMATICS The one track hypothesis claims that anaphora operates on a single track Thus one can measure the distance between pronoun and antecedents in terms of the number of clauses Dubois claimed that there are two tracks or two patterns of anaphora In pattern 1 for subjects subsequent mentions are nonlexical In pattern 2 for objects subsequent mentions remain lexical Peter Tiersma California Supreme Court The Language of Perjury This paper examined a Supreme Court opinion defining a false statement in the law of perjury Bronston us United States Tiersma argued that Grice's maxims operate even in adversarial proceedings and that the court's definition of false statement needs revision PHONOLOGY SEMANTICS Jorge Hankamer UC Santa Cruz Nested Antecedents Natuso Tsujimura Indiana University Japanese Suffixal Accentuation and Lexical Phonology Hankamer cited the following sentence The author identified a class of stem-making suffixes in Japanese which form Bob presented his theory about the intransitive transitive pairs Some of the ECP which is are pretty suffixes show the pattern of recessive suffixes interesting The latter provide an exceptiqn of Halle and Mohanan 1985 and Halle and Vergnaud's where 'which' can be 'his theory about the ECP' 1987 claim that dominant suffixes which are or just 'the ECP' but not the plural entity cyclic must precede recessive suffixes corresponding to the set of the ECP and Bob's theory He used this examples and others to o show how the discourse model approach must be sensitive to syntactic relations Lewis Creary J Mark Gawron John Nerbonne Charles Li UC Santa Barbara Contact Hewlett-Packard Toward a Theory of Locative Induced Tonogensis Reference This paper showed how a Mongolian language This paper proposed a treatment of locatives as Baonan acquired tones after close contact with a tonal language Acoustic analysis of pitch components of arguments to predicates and amplitude of Baonan disyllabic words revealed a system of two tones high and low WORK agent Tom location Li hypothesized that the tonogensis was ON Mass-Ave IN Boston triggered by the infusion of loan words from a tonal language This infusion caused pitch rather than amplitude to become a dominant Toshiyuki Ogihara University of Texas at distinctive feature Austin A Present Tense Embedded Under a Past Tense 3rd Issue 1988 CRYPTOLOG page 20 JPYQK QJPYJPYlGIAb USH ONLY DOCID 4010015 SECOND LANGUAGE ACQUISITION Eileen Kingsley and Ann E Daubney-Davis USC Native Speaker Nonnative Speaker Interactions Variations in Negative Data This paper analyzed six twenty minute interviews of native-nonnative speaker situations in order to identify instances of negative feedback These instances were classified as clarification confirmation reformulation and distress The researchers investigated the relationship between progress and frequency of responses to negative data Results suggest a possible relationship between progress and the amount of negotiated interaction initiated by the learner William Rutherford USC The Contribution of Second Language Acquistion to Learnability Theory Studies in second language acquistion SLA have mainly been studies of developing grammars In contrast theories of learnability or how language can be learned via impoverished data and without negative evidence have only recently been addressed in the SLA literature Rutherford suggested two areas of research in learnability theory the Subset Principle Berwick 1985 and the the Uniqueness Principle Pinker 1986 Examination of these two areas can help explain how learning occurs and thus offer refinement of the theory Monika Farner Macalaster Tutorial Questions in Caretaker Speech This paper examined the language tutoring function of questions of two caretakers German speaking and English speaking of a bilingual child The what-questions of both caretakers were analyzed Results showed that questions of both caretakers increased in complexity over the period of study INVITED LECTURE Charles Ferguson Stanford University The Future of Applied Linguistics An Overview Ferguson stressed the need for identification of problems in the real world requiring linguistic solution and the identification of resources to 3rd Issue 1988 solve these problems He pointed out the emergence of subfields in applied linguistics and how their development meshes with research in other areas For example he cited the effect of phonological theory upon the field of speech therapy PANEL DISCUSSION Writing Effective Reviews Abstracts and Referees' Reports This was a panel discussion sponsored by the committee on the status of women in linguistics Panel members discussed how to write a review for a journal article how to review a grant proposal and how to write an abstract for the LSA conference CENTENNIAL LECTURE in honor of the centennial of Bloomfield's birth Leonard Bloomfield is the father of American linguistics Bloomfield The Man and the Man of Science R H Robins delivered the main lecture on Bloomfield Following Robin's talk a sympoosium of invited speakers Isidore Dyen Murray Emeneau Charles Hockett Henry W Hoenigswald Kenneth L Pike Frank T Siebert William G Moulton and Rulon Wells spoke about Bloomfield's contributions to the field of linguistics AAAL SYMPOSIUM Canadian Immersion Programs Recent Research by Merrill Swan Christina Bratt Paulston Lily Wong-Fillmore Fred Genessee Swain Five different studies were conducted large scale proficiency transfer study age study observation study experimental study The following conclusions were reached a Communicative competence includes discourse sociological and grammatical factors b The teacher needs to point out contrasts between languages and needs to link form with function c Beginning study at a later age is all right because the child needs a firm foundation in one language d Content teaching alone is not effective There is a need for language instruction tied to content teaching Paulston Paulston made a number of comments based on the results above a We- CRYPTOLOG page li'6ft 61i'1i'lelAL tJS 21 6NL f I DOCID 4010015 need to focus upon language teaching not just language learning- Comprehensible input is not enough b The results of the study seem to imply that late immersion is as effective as early immersion One must consider that perhaps early immersion students would have outperformed late ones if they had been taught differently c Methods to discriminate different teaching methods did not reveal any distinct differences in learning because teachers are basically eclectic What is important is the quality of teaching Fillmore Fillmore commented on classroom treatment studies a Evidence from the immersion studies showed that after many years of study children did not achieve nativelike competence It is clear that mere exposure is not enough to insure acquisition Learners need input but they also need to use the language The teachers can have a strong influence upon the degree of mastery b Students got little feedback about how their skills were deficient They had little regular contact with French speakers outside the classroom Genessee Genessee commented on the effect of variation in learner and environmental characteristics upon language mastery a Language achievement is affected by age of arrival and length of residence but length of residence is a better predictor b Earlier study is not necesssarily better Cognitive maturity and Ll proficiency both affect language achievement The International Communications Convention ICC-88 June 1988 Philadelphia Reported by 1 jP13 Optical fiber communications was a mini-theme of ICC-88 About one quarter of the threevolume proceedings con ained information about optical fiber and related telecommunications technology and half of this was from foreign authors Seven complete sessions were committed to the topic of fiber communications This illustrates how important optical fiber is to the future world's communications Within ten years more than 80 percent of the important point-to-point communications will pass over optical fiber links PART I THE CONFERENCE Leading Third World countries viz Brazil and China are actively developing domestic fiber optic production capabilities and intend to catch up with the fiber optic networks and ISDN di tization of the industrial nations by 2000 European and Japanese laboratories and manufacturers and PI'T's are getting ahead of the US in the technology and use of fiber optics The technical development of fiber optic communications is still device-limited The network designers cannot exploit the bandwidth of the fibers because the devices are too expensive too inflexible and too noisy Most of the problems of terrestrial trunk communications have been temporarily solved and the US has installed over a million kilometers of fiber in a competitive scramble for long distance business but the physical fiber is probably unsuitable for future needs The potential demand for fiber to the subscriber is a much larger market perhaps a hundred times larger than the trunk networks but the electronics are expensive and the designers have not pinned down a service they can offer at 150 Mbps or higher which the non-business subscribers will pay for They hope that HDTV will create a demand for the last mile of a total fiber network There has been a lot of activity in the U S laying trunk networks of fiber as more than a dozen competitors sought part of the long distance market after divestiture This has produced overcapacity and mergers and bankruptcies are expected In addition there are severe regulatory and political barriers to development of an all fiber network in the U S as more lawyers have come into the telecommunications field The Asians and Europeans have proceeded more slowly installing much more local fiber than the U S and prolonging the experimental and trial stages while the technology changes This difference was manifested in the sessions where U S authors dominated the fiber trunk sessions while foreign authors had more to say about processing switching and local loops The French have already installed fiber to millions of homes compared to a few hundred residential fiber loops in the U S The foreign 3rd Issue 1988 o CRYPTOLOG o page 22 P L 86-36 FOR OFFICIAL USB ONLY DOCID 4010015 Current Mediterranean Submarine Cables Most of the submarine cables in the Mediterranean are analog systems but some fiber cables have been installed The 1984 NTIA report on submarine cables by Schenck gives maps of the analog cables of that era and also gives technical details There are about 50 analog links which are numbered on the Schenck maps The slide shown by Marone during the oral presentation of paper 2 7 indicated existing and future optical fiber links on many of the same routes as the Schenck maps The analog cables circa 1984 are listed below Presumably many of them will be replaced by fiber or by repeaterless flouride fiber cables within the next ten years Number 21 23 26 34 36 52 53 56 81 90 96 99 100 101 104 107 109 110 113 122 125 129 134 137 146 147 151 152 154 156 157 160 162 174 177 180 184 195 205 206 207 210 215 234 236 237 241 247 249 254 Name Capacity Date Sicily-Malta 36 ch 1955 Ke1ibia-Bou Ficha Tunis 120 1956 Italy-Tunisia 60 1956 Marseille-Bord el Kiffan 60 1957 Italy-Sardinia I 60 1957 Trapani-Cagliari 120 1962 1962 Sicily-Crete 60 Canet Plage-Mers el Kebir 60 1962 Cannes-lIe Rousse 96 1966 Canet plage-Tetouan 1967 96 Marseille-Tel Aviv 96 128 1968 MARTEL Cantanzaro-Lekhaina 480 1969 Italy-Sardinia 2 1969 480 Agrigento-Tripoli 120 1969 Barcelona-Pisa 480 1969 Marseille-Bizerte 96 1969 St Raphael-St Tropez 1970 480 Marseille-Beirut 160 1970 Mediterranean-Atlantic 1480 640 1970 MAT 1 Estepona-Palo Peninsula-Balearic Is 900 1380 1971 PENBAL 1 Pisa-Algiers 480 1972 Marseilles-Bard el Kiffan 480 1972 Cantanzaro-Alexandria 480 1972 Alexandria-Beirut 120 1972 Barcelona-rome 1380 1974 Civitavecchia-Cagliari 1380 1974 France-Greece-Lebanon-Cyprus 480 1974 St Raphael-La Faux 2340 1975 Heraklion-Lagonissi 1380 1975 TELPAL Tel Aviv-Palo 1380 1975 ANNIBAL Perpig -Bizerte 640 1975 Algeria-Spain 480 1975 MARPAL Marseilles-Plao 2580 1976 Italy-Turkey 480 1976 Rome-Palermo l800 2tv 1977 Marseil1e-Bastia 2340 1977 PENBAL 2 3900 1977 AMITIE' France-Morroco 2340 1978 Tripoli-Benghazi 900 2tv 1979 Genoa-Sassari 3600 1979 La Seyne-Tripoli 640 1979 BARGEN Barcelona-Genoa 4140 1979 France-Algeria 4 2580 1980 Greece-Syria PALMYRA 480 1981 Greece-Cyprus 2 APOLLO 1380 1981 France-Greece 2 ARTEMIS 2580 1981 Sardinia-Sicily 2 3900 1982 Juan Ie Pins-Cagnes sur Mer France-France fiber optic 340 mbps 1982 France-Tunisia 3 2580 1983 Greece-Egypt ALEXANDROS 624 1983 3rd Issue 1988 o CRYPTOLOG o page 23 FOR OFFI6b L USE ONLY Constr contract sub cables cit bpo cit stc stc scI cit cit cit excofina scI stc stc stc cit submarcom cit stc C w stc cit stc cit stc stc submarcom submarcom stc stc submarcom submarcom submarcom stc stc submarcom stc submarcom NEe stc submarcom stc submarcom submarcom stc submarcom stc submarcom submarcom att I DOCID 4010015 PIT's can develop fiber nets as part of an integrated telecommunications plan This takes longer but costs less and probably avoids the rapid obsolescence and waste of the American efforts One point of interest was the development of flouride based fiber for repeaterless submarine cable links in the Mediterranean by the Italian CSELT They expect that transmission at long wavelength will allow 300 Km repeaterless links at 2 4 Gbps by 1992 and even longer spans by 2000 Another point was the prevalence of damage to the fiber trunks from backhoes and other effects which is spurring the design of survivable networks There is a plan for a global synchronous optical network SONET which will use TDM multiplexing to carry 150 Mbps services at line rates of 2 4 Gbps 16 x 150 2400 However there are critical timing problems in a synchronous net which are exacerbated by the US divestiture that has different interconnecting networks driven by uncoordinated clocks Although there is intensive research and progress the manufacturing of certain types of fiber and components is still difficult and the development and operation of optical fiber networks is still a jungle of problems Costs applications markets and integrating the wideband capacity of the fibers into the existing narrowband copper and radio nets are creating uncertainties and delays in the actual deployments of fiber and the development of machines to interface fiber with non-fiber nets In the meantime the technology is spreading all over the world AN OVERVIEW OF THE SESSIONS Fourteen of the 54 sessions dealt entirely or partly with optical fiber Western Mediterranean Sea 3rd Issue 1988 o CRYPTOLOG o page 24 FOR OFFICIAL USI3 ONLY I DOCID 4010015 Session 1 on optical switching presented different schemes for switching of trunks MANs LANs and subscriber loops The designers are trying to develop an all-optical net All of the papers were foreign Session 2 on long-haul fiber systems was dominated by Americans The competitive race to build long distance fiber trunks by SPRINT WILTEL and others has required elaborate testing and installation procedures because the new carriers did not have organic resources to do the installation and splicing Undersea fiber cables will have wetmux switching and repeaterless flouride glass links because it will not prevent polarization shifts that degrade coherent transmission Session 11 on digital cross-connects systems described the machines for interfacing between fiber trunks and local networks particularly fiber hubbing nets The existing asynchronous digital networks using radio and cable at electronic speeds must connect into high speed synchronous optical nets SONET Session 19 on simulation of lightwave systems covered mathematical models of fiber lasers waveforms and subnetworks Session 28 on fiber access to the home current work in the critical last described Session 10 on high speed transmission was mile of fiber installation Costs are a major mostly foreign papers Foreign laboratories have demonstrated 10 Ghz linear amplifiers and factor The Japanese and French are trying to receivers and 700 Km repeaterless spans Fast get an economical transition from current services to future wideband services They have transmission is device-bound and although fewer regulatory barriers to the PTT's more than a million kilometers of fiber have delivering CATV but still have to deal with been installed it is probably the wrong kind Eastern Mediterranean Sea 3rd Issue 1988 CRYPTOLOG peR eFFIClIAL us page 25 ONbY I DOCID 4010015 established interests Some of the architectures create an ether in which all the traffic reaches all the subscribers while other designs select traffic flow at central switches Combinations of fiber and coaxial cable are use to reduce the number of fiber splices and taps Analog transmission of TV is much cheaper than digital coding and decoding taps and bridges Synchronization of many independent local and long distance digital fiber nets will be very tricky maintaining fiber which they think necessary for coherent systems Coherent lightwave communications CLC can be used to carry microwave compatible transmissions based on existing microwave systems Optical amplifiers will give 20-30 dB gains AT T thinks a 4 5 terabit sec star net can be built with currently available components The decisions to install monomode fiber that does not preserve polarization and adopt TDM multiplexing will dominate switching and CLC Session 29 on broadband systems showed GTE's preference for circuit over packet switching Siemens considers 150-600 Mbps services to subscribers over fiber loops Fujitsu prefers Asynchronous Transfer Machines ATM for interface between narrowband ISDN at 64 Kbps and broadband ISDN at 150 Mbps Memory matrix TDM switches allow 8 streams of 150 Mbps to be combined into a 1 2 Gbps signal in a switch with a throughput of 8 x 1 2 9 6 Gbps currently demonstrated The Japanese like Transport Processors TP for the expected bursty nature of multi-gigabit broadband services Session 46 on lightwave local networks showed a number of schemes to operate nets at terabit rates One academicians's scheme used pulse position addressing combined with spread spectrum transmission over fiber Multi-fiber rings allowed many different architectures to be expressed over the same robust physical net The Japanese proposed fiber LAN based on combining FDDI access with CSMA CD transmission Session 37 on coherent lightwave transmission was dominated by U S papers The Japanese noted problems in manufacture of polarizationAnswers to CRYPI'O- LOG PUZZLE 2nd Issue 1988 1 2 3 4 S 6 7 8 9 10 II 12 13 14 15 16 17 18 Easy as falling off a log Psychological Logjam Lincoln logs Loge Logrolling Logy Ship's log or logbook Catalog Sawing logs Slept like a log Tagalog At logger-heads Backlog Sat like a bump on a log Yule log Analog computer Log cabin Session 47 on broadband ISDN and packet switching was concerned in part with how to extend broadband ISDN services to residential customers TV and HDTV are seen as the main traffic to non-business BISDN customers Broadband ISDN will require a global SONET synchronous optical net and ATM asynchronou8 transfer machines for interface to the existing asynchronous digital networks 00 00 00 In general the cost of the electronics is too high for subscriber optical loops Europeans and Asians are able to pursue systematic development of integrated optical networks without the regulatory problems and fragmentation that affects the U S telecom environment The U S has deployed a lot of yesterday's technology for todays short-term market share while the foreigners are proceeding more slowly but on a broader front to get the markets the costs the technology and the customer demands coordinated in a profitable way In some areas e g subscriber services they are well ahead of the U S This may give them a long-term advantage even inside the U S market 3rd Issue 1988 o CRYPTOLOG o page 26 li'6ft 6li'li'ICIAL USE ONLY DocrD 4010015 eONFI8I'3N'fIAL PART II SIGINT IMPLICATIONS OF TRENDS IN FIBER OPTICS COMMUNICATIONS U Most of the world's long distance point-topoint communications will pass over optical fiber trunks within fifteen years The capacity of the fiber trunks and the low cost of transmission tends to concentrate enormous amounts of traffic from thousands of low speed tributary systems onto a single fiber U The impact of fiber is so significant that even COMSAT is installing fiber Fiber is already displacing traffic from satellite and microwave links For broadband digital services it has no equal Bit rates above 3 Gbps are already in service and rates above 50 Gbps will soon be in use on single fibers At the same time local nets will be carrying i terabits of traffic and on many local nets all the bits will be available to all the subscribers Analog transmission of microwave signals will also occur on fiber probably as part of an analog-digital format Improvement l in glasses and optical components will allow ooth high bit rates and long repeaterless span Infrared wavelengths at 3 4 microns wilfbe used to take advantage of low attenuation in new glasses On DOG The TRANSCRIPTION ME O fOW MEOW lB' MElHYI MEOW MEOW IPURRSJI MEOW MEOW MEO ' IG pnss MEOW Lighter g g 0 MEOW MEOW MOEW MEOW MEOW MEOW MEOI ' MEOW IfIIISS MEOW MEOW MEOW Side EO 1 4 e P L 86-36 by Charles Ralya courtesy of' Vox Topics 3rd Issue 1988 o CRYPTOLOG o page 27 CONFIf'l'3fifTIAL IIARBLE YIA eOftHN'P CIIMHf8LS OnLY DOCID I 4010015 Call for Speakers WHEN YOU WERE IN SCHOOL DID YOU EVER WONDER WHY DO WE HAVE TO SOLVE THESE DARN LOGARITHMS WHY DO WE HAVE TO KEEP PROVING THAT TRIANGLES ARE CONGRUENT WHY DO WE HAVE TO LEARN MATH I'M NEVER GOING TO USE IT STUDENTS ARE STILL ASKING THESE QUESTIONS THE LOSS OF STUDENT INTEREST IN MATH IN PAST YEARS IS THREATENING THE FUTURE OF U S MATH AND TECHNOLOGY THENSA MATH SPEAKERS BUREAU HAS BEEN ESTABLISHED BY THE MATH PANEL WITH THE GOAL OF INCREASING YOUNG STUDENTS' INTEREST IN MATHEMATICS You can help answer the WHY questions by visiting Elementary Junior High or High School math classes in the State of Maryland as a Guest Spe3ker Stimulate students' minds with discussions of o the beauty of mathematics in nature o elementary cryptology o how to solve logic puzzles o careers in math I math in careers o any topic that you would like to presentwe have a library of topics to give you ideas You DON'T HAVE TO BE A MATHEMATICIAN TO BE A MATH SPEAKER For more infonnation contaci coordinator NSA Math Speakers Bureau G423 963-6423s 688-5361b lithe math community suffers the National Security Agency suffers -i Chainnan NSA Math Panel 3rd Issue 1988 o CRYPTOLOG FOa OFFICIAb y page 28 OHbJPY DOCID 4010015 L LEFTOVERS byl _ This puzzle is UNCLASSIFIED Each of Ibe clues below conrains defmitions for four 4-leuer words Removing one letter from each of Ibe fll st Ibree 4-leuer words and combining Ibe resulting 3-leuer pieces in Ibe order given will form a new 9-lettee word 11Ie letters you removed also taken in order will spen a 3-letter word You can now add a letter to some position of this 3-letter word to fonn Ibe answer to Ibe fourth defInition The letter you added is Ibe leftover The seven leftovers in order spell he solution word Call me with the correct solution at 1351 s and I'll print your name in Ibe next issue Ex Bony growth SPUR Ice arena Give temporarily Singe RINK LEND BURN B Remove Ibe U from SPUR the R from RINK and Ibe N from LEND Combine Ibe 3-leuer pieces to and adding Ibe letter B fonns Ibe fonn be wad SPR INK LED The removed letlQS spell Ibe word word BURN the answer to the fourth definition The leftover is Ibe B URN 1 foot or an ear Something on a College figure Dispatch Pour 2 Beta Orionis e g Bucket handle Sympathy Incline 3 Enos' grandfather Egyptian goddess Wood's partner Join 4 Way out Contact On stage Streamlet S Imitator Crossing Small lake Assistant 6 Concrete Smoker Feathers' prerequisites Fuzz 7 Road hazards Average Recounting A fuel Courtesy of CRIB 3rd Issue 1988 CRYPTOLOG page FQR QFFIGIAb l JS 29 QNbJPY 86-36 I DOCID 4010015 -feP-SEEREf CJ il ' - -_ _---IL -_ _ Tilli QQ M T Q T 1 8 CODEWORD MA'I'ERIAb -f6P-SEERE J-NOT RELEASABLE TO CONTRACTORS 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