86-36

fil l5lDWfill5 fil II Wl5Ul fill5 f 00UlW"l1aJlIJfil

P.L. 86-36
I

December 1981
EXERCISE SUPPORT (U) ........ ..
THE 1981 URSI xx GENERAL ASSEMBLY (U) ......
SLEEP WELL I YOUR SDO IS ON DUTY (U) ...... E. Leigh Sawyer . 8
tIDE: A BRIEF HISTORY (U) .. 1 I. ... . ... 10
NSA-CROSTIC NO. 36 (U) .... David H. Williams '.' '.' . . .. 14
IN PURSUIT OF: FASTER HORSES, YOUNGER WOMEN,
OLDER WHISKEY, AND MORE MONEY (U) .
TEXTA: WHAT IS IT? WHERE IS IT GOING? (U) .....
PLATFORM: HOW DID YOU SAY THAT WORKS? (U) T--
REVIEW: IN THE NAME OF EFFICIENCY (U) ..
L...- ----I
DOCIDl:-- 40111945 ./
P.6
'fHIS B6CtffltIHN'f C6N'f2\INS C6BHW6RB
B N8A/S88M 123 2
REVIEW eN 18 Bee e811
Declassified and Approved for Release by NSA on '10-'1.2-.20'1.2 pursuant to E.O. '135.26.
vl DR Case # 54778
DOCID:
Publishld by Pl, Techniques and Standards,
for the Personnel of Operations
VOL VIII No. 12
PUBLISHER
DECEMBER 1981
EDITORIAL (,oJj
It is sometimes puzzling wny some people
turn out to be good at this analysis business,
while others who are evidently just as well
educated never seem to get the knack of it. I
have, of course, some theories about this (or
I wouldn't have brought the subject up), and 1
wonder what you the readers think about this.
BOARD OF EDITORS
Editor-in-chief1 1<7119S)

Production ...... 1 l3369S)
COllectionI
cryptanalysis .. 1
CryptolinguistiCS.. 1
Information Science. I [03034s)
Language ...........1
Machine Support. I lt5084S)
Mathematics .....
Puzzles ............. David H. Williams (1103s)
Special Research ....... Vera R. Filby (T!1'\19s)
Traffic Analysis .......Don Taurone
For individual subscriptions
send
name and organizational designator
tOr CRYPTOLOG.
r33
69S or call
\\
To submit 'articles or letters
via PLATFORM, address to
cryptolg at bar1c05
(note: no '0' in 'log')
One notion I keep coming back to, is that
very few discoveries are ever made by means of
the so-called "scientific method", and that it
is only when one comes to the point of having
to describe the discovery to others that the
"scientific method" is used, as an orderly way
of laying out facts and their connecting argu-
ments.
Over the years, it seems to me that the
great bulk of the genuine analytic discoveries
were made by a relatively few people. Some of
these people, perhaps most, were not really
"scientific method" people. If they had a
"method" there was often a faint smell of
magic about their description of it. Some-
times, they would describe their discovery as
an accident, as pure serendipity.
1 no longer believe in the serendipity
explanation. It simply doesn't fit the cir-
cumstances. The accidents should have hap-
pened to a larger number of people, not just
to a small number of "serendipity-prone"
analysts. I think it has to do with the way
these people looked at the world, the way they
perceived events around them.
Based upon my contact with a few of them, I
believe that some of these serendipity-prones
looked at the world around them in terms of
a before,
an after, and
the event connecting them.
Given a certain "before" and "after" that
were .Il2i the same, something "happened" at the
connecting event. I think it may have been
this way of looking at the events around them,
that drew the attention of these serendipity-
prones to the sites of their discoveries.
Something did.
What do you think?
P.L. 86-36
DO-e-ILl: 4 O...... 1...... .................. .
. L. 86-36
!
Exercise Support (LJ)
.....,.,r
xx
C8:J
N
SA!C5S, in accordance. with USSID 4,
supports U.S. military exercises
where CXl'UNT 1s/ required. V42
(Current AppIlcations Division)
--1'r(C""'=C"""'CV",Tj-levies perSQnnel to serve as SIGINT
Reporter/Analysts the exercises. Since
January, 1981, A6/(Technical Support) has par-
ticipated in the program. '!he author would
1 j ke to eYD.[ess his appreciation to Michael F.
---.P1ief V42l, for his ccmnents on the
section concerning the recent exercise.
Introduction to ]!:xercise SIGINT (U)
(U) All large military exercises are con-
ducted on the basis of a scenario depicting
-sane imaginary war situation, both to give the
participants experience in coping with situa-
tions not encountered in peacetime and to
evaluate the capabilities of coornanders,
staffs, troops, systems, and equipnent to cope
with such situations.
(U) In the course of the exercise, partici-
pants are given preplanned information about
imaginary events in the scenario - what the
enemy is doing, what casualties friendly
forces have suffered, et,c. - and are expected
to respond to these "events" with orders,
plans, and actions. Exercise controllers
then assess the results of the players'
actions, give the players appropriate feedback
through simulated intelligence and operational
reports, and devise subseqUent problems for
the players.
(U) Because players' actions are often
unpredictable, the course of events can devi-
ate considerably from the original script.
Controllers are expected to be able to handle
this "free play" and still keep to the key
themes of the scenario.
(U) MaXimUIII realism is sought, but realism
will always be sacrificed to accomplish exer-
cise objectives. For example, enemy capabili-
ties are adjusted as required to provide the
desired amount of challenge to the players,
even to the point of wild implausibility if
necessary. The war cannot be allowed to be
won or lost desperate battles
must continue right up to the end of the exer-
cise.
(U) Chaos and confusion are unavoidable
features of real warfare. '!hese qualities are
also characteristic in the management of
intricate exercise scenarios. '!he planned
breakdowns and disruptions designed to test
the players are canpounded by unplanned break-
downs and disruptions inevitably suffered
among the controllers and referees.
Dec 81 * CRYProIDG * Page 1
./
EO 1.4. (c)
P.L. 86-36
IWIBhl3 VIA SSHIIIIf 8IWlNElhS SflhY
DOCID: 4011945
A Recent Exercise (U)
CPr IIthen returned to
NSA/CSS and began the SIGINT,
personnel and logistic support for the exer-
cise. Included in these support efforts were
airlift requirements for NSA/CSS
sUpPo{t personnel and arranging for ac;cOl1llOOda-
tions aootransportation at the exercise. site.
He also reqUested the necessary personnel sup-
port from A Gtoup and communications support
fram oor.
(U) The purpose of this JRX was to provide
training for participating commanders,. staffs
and forces in joint air/ground operatioQs
involving air, armor, and mechanized forces,
............. 1
IU) noon notification of the JRX, captain
I lof V421 (Exercise Support
tlrancn) was aeslgnated NSA/CSS project off-
icer. He attended the initial planning
conference at REOCGi Hq (McDill AFB) where all
of the intelligence players were brought
together and the strategic scenario was given.
exercise scenario
P.L. 86-36
EO 1. 4 . (c)
1/// .L . 86 - 3 6
Dec 81 * CRYPTOLOG * Page 2
'Ii:' 8BHHI'f 81h'rlUl'CbS BIlbY
SSSRS'f
(C-CCO) The Joint Exercise Control Group
(JECG) was the Exercise Controller, and as
such interjected stimuli into the exercise to
which the players responded. The SIGINT Sup-
port Liaison Staff (V421) was part of the
JECG.
(c-ceot The Cryptologic Support Group (CSG)
and Intelligence Directorate (J2) were co-
located and were provided SIGINT inputs via
the Hobile Cryptologic Support Facility
(HCSF), which is a GH motor home that houses
the most modern computer and communications
equipment. The HCSF belong to NSA/CSS.
I
P.L. 86-36
;;0 1. 4 . (c)
't'G.l CPT
II
said that the .After Action
Report submitto the JCS
would probably conclude that the JRX achieved
its objectives primarily because logistics and
air defense objectives were met.
Ihowever, doubts the value
of the exercise as a realistic simulation for
the following reasons:
-rc-eee1-Upon conclusion of a JRX, a brief
I0:' L. 8
S ou nave participated, but, in
only the NSA/CSS personnel attended.
1
xxx

1
xxx

1
(U) The CSG is a group provided by NSA/CSS
to facilitate SIGINT Support to a unified or
specified command, joint task force commander
or other commanders.
(C-CCO) The Consolidated SIGINT Support
1_--1
The SIGINT Support Staff (SSS) gen-
Grated the SIGINT support for the exercise.
(U) As to the success of SSS in support of
the. exercise , one can only conclude that exer-
cise support objectives were partially met.
/i
Dec 81 CRYPTOLOG Page 3
SSSREl'i'
IIAP/BY VIA 88f11IN1' 8HAPIPIIIlS 8NM
. -_.-._-- ..._------
...
...
...
...
...
IL....--- I
(U) Transcribers are left with a more
complete view of current operations and a more
complete understanding of the assorted and
sanitized reports that in the intelli-
gence community.
There was gener.l agreement among
the participating transcribers on two points:
.. Exercise scenarios should De provided
with sufficient time for perusal;
.. Personnel ishould" be carefully selected to
insure that the individual has the proper
background.
Transcriber Reactions
to Program Participation (U)
(U) Although the opinions varied among the
transcribers participating in the various
exercises, there were many areas of general
agreement. Most of the transcribers thought
the experience worthwhile in that they saw how
the Agency fits into the SIGINT community,
i.e., into "the big picture." Most thought
they had gained from the experience, either
professionally by a greater understanding of
the SIGINT system and an increase in target
knowledge, or financially by the overtime.
(U) At the time of A6's entry into the pro-
gram, transcribers suffered from poorly organ-
ized pre-exercise briefings, little working
aid familiarization, instructions that were
wrong or incomplete or too rapidly given, and
disorganized source materials. One tran-
scriber noted that the standard was to be
uninformed. The various staffs, however, are
correcting these shortcomings on a continuing
basis.
Otherwise, the individual maybe
with/a mass/of incomprehensible data
no time to decipher it. Some
knowledge lis requisite.
-4" ... r
inundated
and given
military
Dec 81 CRYPTOLOG Page 4
'll'" SQUHR:
eellf'IBBli'l'IAh
recently attended the 1981 URSI
his report offers some unset-
tling projections about the problems that
SIGINT will face over the next one or two
decades. This is an extract from the In-
troduction and Conclusions of that report .
The 1981 DRSI
:xx General Assembly lJ;
I
I
1
j
I

I
. L. 86-36
h, 1981 URSI (Int,rnationaI Radio
Science Union) meeting was held in
Washington, D.C. 10-19 August, 1981.
The international meeting is held
(U) every three years, and this was the
first time in several decades that it occurred
in the U.S.
(U) Because the meeting is prestigious, the
authors and national radio science organiza-
tions make an effort to publish significant
work. About 500 technical papers were given,
covering almost every aspect of radio theory,
including optical fiber, computer design,
instrumentation, remote sensing, and biologi-
cal effects of radio on humans, as well as the
more conventional areas of propagation, noise,
radio astronomy, microwave power, satellites,
and telecommunications.
(U) The authorship, as well as the atten-
dance, was very international. The meeting
organizers reported that 1056 people
registered, from 38 different countries.
About half of the audience was from the USA.
Japan, France, Germany, and the UK also had
large contingents of 50 to 70 attendees. Six
people came from the USSR, including some
authors. 269 of the papers had foreign
authors, although in some technical areas,
nearly all the authors were from the USA.
are some interesting paral-
lels between SlGINT and radio science, viz:
1. Both activities deal with radio in a
very broad way.
2. Both are concerned with quasi-
repetitious phenomena, which they can-
not control, as well as with unique or
very random phenomena which give one-
time capture opportunities.
3. Both have to develop unique apparatus
and unique processes to obtain data
and to extract information from these
data, so they are both concerned with
device engineering and measurement
techniques.
4. The volumes and bandwidths of data are
often very large, and the "explana-
tions" tentative.
I (C-eeS) In general radio sCienti:ts J
understanding the physica phenomena of rao1o,
communications, and instrumentation. Their
discoveries and measurements have, over many
years, opened up new areas of the for
radio applications, and improved instruments.
They have also shown basic physical limita-
tions to uses of radio, e.g., tile effects of
vec 81 CRYPTOLOG Page 5
EO 1.4. (c)
P.L. 86-36
SStlf'IQStITI A b
IhltllBMl IfIA 8SIIIU'l' 81h1tlllll!LhS SUb
DOCID: 40119lf5
SEleREl'f
(8 ese) The telecommunications planners and
designers in many countries pay attention to
the findings of radio science in the operation
and particularly the develo ent of their s s-
tems and e ui ent.
water vapor on short radio
bulence on laser beams,
noise on receiving systems.
waves, air tur-
and the effects of
Analysis (U)
(U) Some of the major developments in com-
munications have stemmed from advances in
materials and the invention of devices. The
electric telegraph of 150 years ago resulted
from the purification of copper so that cir-
cuit losses were reduced to low enough levels
to make generators, relay windings, and tele-
graph lines feasible. The semiconductor
explosion came from improvements in the
materials of semi-metals, and the current work
in improving glass and optical devices seems
to be setting the stage for a major revolution
in switched communications.
(8 eee) One thread of technical information
from the URSI meeting will illustrate this
point. Developments in optics have shown that
great improvements in cost, performance, and
lifespan of semiconductor lasers are
with lifetimes of 100,000 hours confidently
predicted, and 'million hour lifetime thought
possible. The fibers themselves are getting
better with bandwidths of several and for
certain new fibers, hundreds of GHz capacity
are expected. From this progress, the CCITT
(International Consultative on Tele-
phone and Telegraph) is now develqping stan-
dards for worldwide compatibility between the
parameters of all public carrier optical
fibers, so the optical fiber can inter-
face. According to a French consultant, the
Europeans look upon satellites/as a temporary
measure for regional communications (TELECOM
1) and will shift all main line transmission
to optical fiber links across Europe as fast
as they can lay in the trunks, with the satel-
lites reserved for mobile and other light ser-
vices. At the same time, studies of the phy-
sical characteristics of optical components
indicate to other Europeans researchers that
the l2Qal networks cannot use optical fiber to
carry 50 CATV signals in a bus, so they expect
to go to a switched optical fiber network to
replace the eXisting copper wire local plant
with an individual fiber from a switch to each
subscriber. This network would be expected to
serve for 50 ears after installation.
(U) A recent survey of telecommunications
in the Economist, 22 August 19b1, notes the
progress in optical fiber systems which
surprised even AT&T and the BPO (British Post
Office). Over the next decade, the world will
spend 640 billion dollars on telecommunica-
tions equipment (according to an A. D. Little
study), and radio will be a significant part
of that. As optical'fiber trunks take over
the main line transmission loads, and even
spread into the local networks, the radio fre-
quencies will be applied to mobile radio,
satellite service (expected to exceed 700,000
Intelsat circuits by 2000, plus even greater
domestic satellite capacity), and many ser-
vices where wires or l'ight guides are impossi-
ble or impractical.
(U) One of the notable features of the URSI
is the close interaction between .devices,
materials, and radio.technology. Remote sens-
ing depends on microcircuitry, super comput-
ers, models and devices to compensate for
atmospheric distortion, propagation theory,
antennas, and so on. Spread spectrum or radar
signals to combat radio interference, noise or
propagation effects depend on 5uperspeed
equalizing digital processors, electro-
acoustic analog convolvers (waveform compar-
ers), and high bit rate key generators, etc.
Even efficient and accurate television
transmission depends on sophicsticated signal
processing technology, e.g . , SAW PAL filters
that are used in hundreds of thousands of TV
receivers to overcome adjacent channel
interference.
Conclusions (U)
(U) Developments in radio science bring
about a large number of gradual improvements
in the big telecommunications systems, as. well
as initiating some radical changes. Both
kinds of change accumulate to bring complete
I
!
EO 1.4. (c)
P.L. 86-36
Dec 81 CRYPTOLOG Page 6
S8SR8'l'
IWIBhS VHf 8eUHR SIIMaIST.S em.v
-DOCI D: ............ f
1. 4. (c)
.L. 86-36
transformations in the way telecommunications
operate and affect things.
(U) The main impact of radio and optic
developments will be in transmission, where a
generation of technology lasts about 15 years.
The effect on switching, where technology life
cycles are about 30 years, will be delayed
until the 1990's, after the current commit-
ments to digital electronic switching are ful-
filled.
Dec 81 CRYPTOLOG Page 1
j
lWIBbB nit S811111; SnMI!IIH.s 9N1.
DOCID: 4011945
UNCLASSIFIED
by E. Leigh Sawyer, 14
SLEEP WELL!
YOUR IDG
IS ON DUTY!
#
seem to detect a growing trend for
people to rummage around in their
cryptologic attics to describe cer-
tain events or occasions taking
(U) place in the olden days. Doing a
little rummaging on my own, I recalled a
long-abandoned function once carried out by
company grade officers identified, as they
popped up periodically on the master roster,
as "SECURITY DUTY OFFICERS" (SDO). This sys-
tem was in its heyday in the early 1950's. At
that time, AFSA (soon to be NSA) was split
between the Naval Security Station at Nebraska
Avenue and Arlington Hall Station. As a
digression, this split had its interesting
features too -- like the time I drove my car
from NSS to AHS during the day and took the
shuttle bus back. At quitting time, I natur-
ally couldn't find my car in the NSS parking
lot and was at the point of reporting a stolen
car case to SEC when I realized what I had
done. I managed to catch the last shuttle
back to AHS by the skin of my teeth. Ah,
those were the daysl Well anyhOW, back to
this SOO system. It was used at NSS (whether
it was used at AHS escapes me). In any case,
it was a so-called "sleep watch." For this
purpose, a cot was located in SEC spaces so
that the SOO could catch some sleep between
his late evening and early morning ,rounds of
all the AFSA spaces.
(U) A word or two about the cot might be in
order. The mattress was obviously not config-
ured for sleeping purposes. What it was
stuffed with must remain somewhat problemati-
cal,' but I suspect it was a mixture of corn
cobs and pine cones. I wonder if somewhere in
the archives there still exist the logs main-
tained by the SOO's. References to that mat-
tress were rife in these logs, e.g. , "after
the worst night of my life", "started my morn-
ing round with every bone in my body screaming
agonized protests", "millions for a new build-
ing at Meade; why can't SEC provide a decent
mattress?" and "even my teeth hurt."
(U) It should not require too much imagina-
tion to determine that periodic one night
stands of this sort were mighty boring. So
Dec 81 * CRYPTOLOG * Page 8
UNCLASSIFIED
DOCID:
I
eoltPf!'!ltUAL
what do you do to lessen the boredom? Drink
coffee, of course. Unfortunately, there was
not a hint of either coffee or coffee-making
paraphernalia discernible in the SEC spaces.
This led to a preliminary round of visits to
adjacent spaces to locate a coffee mess, relo-
cate the pot and can of coffee to SEC spaces,
and return them carefully in the early morning
hours. After all, "nothing's too good for the
boys in the Service." This system was not
without its pitfalls. Like the Navy lieu-
tenant, whose name is no longer retrievable
from my data base, who had a memory lapse and
couldn't remember the office from which he
made the "borrow." He handled the situation
neatly, however, with the following entry in
the log: "Coffee pot and can of coffee found
adrift in Building" 18."
(U) It "was somewhat rankling that SEC,
which conceived this Security Duty Officer
concept in the first place, couldn't provide a
little lousy coffee for us. I recall roaming
around the SEC spaces one evening surveying
all the possible places where they might hide
their coffee and equipment. The survey nar-
rowed down to one small cabinet locked with a
brand new shiny Sargent and Greenleaf combina-
tion padlock. The cabinet had undeniable cof-
fee stains on top and was so deorepit that
forced entry would have probably taken about
15 seconds. I reasoned that a rickety cabinet
of this sort oertainly wouldn't be used for
anything classified. So why the formidable
padlock? "You don't suppose", I mused to
myself. I then proceeded to dial 10-20-30,
the factory-set combination. EUREKA I It
opened. When I opened the door, I beheld
a complete coffee mess. It was not exactly
as though I was discovering the tomb of Tutan-
khamen, but the sensation was somewhat the
same. I shared this revelation with a few of
my friends who were also obliged to stand SDO
duty. But, alas, the SEC coffee mess vice
president must have discerned that the coffee
level was dropping far faster than it should.
Accordingly, it was not too long after my ini-
tial discovery that 10-20-30 no longer worked.
(U) The real psychic bennies for the secu-
rity watch types were in the form of finding
classified materials "adrift." As a result,
there was no doubt in my mind that we were
looked upon as pests by the various organiza-
tions making up our beat. For that reason,
access doors to the various operational spaces
were generally kept locked to keep us (and
incidentally others) out. This meant that the
Security Duty Officer most often was limited
to walking along murky corridors in the vari-
ous buildings used by AFSA in the NSS com-
pound. However, on one occasion, circa 1952,
I found one of the doors open to an RID area.
Oh man, the fox was really in the hen house I
After I had spent a good deal of time going
through every nook and cranny where something
classified might be lurking, I finally was
rewarded -- a classified manual (CONFIDEN-
TIAL). I recall that I thumbed through it
and, in retrospect, imagine if it had been
tossed over the fence into the Russian Embassy
compound, they would probably have thrown it
back out. However, it was marked CONFIDENTIAL
and that made it fair game. As best as I
could determine, the owner of the bookcase was
identifiable by the name plate on the nearby
desk. So I wrote it up dutifully noting the
name of the responsible person:
DR. LOUIS W. TORDELLA
tet As luck would have it, Dr. Lou was soon
after reassigned to the Plans and Policy Divi-
sion, where I worked, to spearhead a highly
innovative experiment called "Third Party". I
confessed to him that I was the culprit who
gigged him and expressed the hope that I had
not set him back in his career in some
fashion.
I
(U) So much for the SDO system. It
the same way as smudged carbon copies of
SUMs, A and B buildings, AFSA-062 "and
063, and red phones.
Ou sont les neiges d'antan?
went
TECH-
AFSA-
Dec 81 * CRYPTOLOG * Page 9
8SNFiBElN'flAb
DOCID: 4011945
EO 1.4. (c)
P.L. 86-36
SBSRS'!'
P.L. 86-36
by11.....-- 1
TIDE:
A Brief History
Author's Note:
Too often in our business, a project's history
i8 written in a coLd, hard, bureaucratic styLe. I
In this paper, I attempt to describe in some-
what hUman terms the story of one of this
agency.'s more successfuL, aLbeit ve:r:atious,
computer systems.
Dec 81 CRYPTOLOG Page 10

T343
SBSRBi
!
i
,I"
I
OCI D: 4 -..I.1_-.!-I
F.L.86-36
S8eRS'f
Dec 81 Page 11
iiSRBT 1I:.N9J:.8 'fH SettlN'f SII/.NN8J:.S 8NJ:.!
DOCID: 4011945
////
EO 1.4. (c)
P.L. 86-36
:3eeflEi'f
Dec 81 CRYPTOLOG Page 12
:3eefle, IIItMBbB nA SertiNi SHhNHSb8 QNbY
--=- -=-=- -----
-nOCID. I
1

/
reconditioned, and, if lucky, enter a retire-
ment of loving care by high school or college
students.
(U) To sum up our experience - this
Agency's use of TIDE has not been an easy one.
Many of those associated with TIDE, however,
believe that the processing crises and sleep-
less nights were worth the achievements this
system has somehow performed. It is believed
that TIDE produced far beyond its expecta-
tions, and many believe the feats it per-
formed, and the people who have made it possi-
ble to simply "maintain" for such a period of
time, should be commended.
(U) However, no matter what its history,
TIDE was simply a little imaginative software,
two machines, and an assortment of peripheral
equipment; When it failed to respond to a
crisis or an analyst was unable to retrieve
important intelligence, it simply became a
use:).ess tangle of wires.
CRYPTIC CROSSWORD SOLUTION, November 1981
EO 1.4. (c)
(par + s.+ dips) 86-36
Relief at Last (u)
(U) Thereafter in rapid succession over the
summer of 1981, more TIDE processing systems
were accepted by PREFACE tbereby providing
additional TIDE relief and yes, some well
deserved rest.
(U) Although some disruptions and minimiza-
tions still occur on TIDE, they are infrequent
(in comparison to previous events) and are
primarily caused by hardware and/or software
failures - not solely loading demands. TIDE
is looking forward to retirement.
The Future (U)
1.
5.
9.
10.
12.
13.
14.
17.
20.
23.
24.
25.
28.
29.
30.
3Y.
PARSNIPS
ARMPIT
COUPLETS (double definition)
SONORA (son +.or + a)
EATS (anag.)
STOAT i&
PUNT (to:Q. .:lm.t.ied)
DEMONSTRAToR (demon's tractor - c)
MELODRAMATIC (anag.)
RARE (double definition)
SEALS (less + a; anag.)
FUZZ (double definition)
LEGEND (leg + end)
CLERICAL (cleric + la reversed)
ANTHEM (ant + hem)
ANACONDA (Dana anag. + a + con)
(U) Once TIDE is finally relieved of its
remaining terminal responsibilities (e.g. high
speed printers,. CRTs, etc.) a full decade
since its creation, it can be unplugged.
Because the soul of any computer system is its
software, what was once called TIDE will
remain only two antiquated 1965-vintage UNIVAC
494s. These machines will be returned to
their Mid-West Minneapolis birthplace, be
1. PICKET (double definition)
2. ROUSTS (r + ousts)
3. NILE (anag.)
4. PETITION ANEW (anag.)
6. MOOR (reverse spelling)
7. PRODUCTS (pro + ducts)
8. TRAITORS (anag.)
11. MARSHALL PLAN (marshal + I + scheme)
15. READS (pun for reeds)
16. DRAIN (D + rain)
18. UMBRELLA (anag.)
19. ALL RIGHT (everyone + not left)
21. VULCANIZE (Vulcan + ize)
22. AZALEA (as + a + lea)
26. ONCE
27. ERIC
Dec 81 CRYPTOLOG Page 13
11M/BY "IN 89"11/'/'1' 811...'IHllhS 9HIiY
SI!SM'f
DOCID: 4011945 UNCLASSIFIED
NSA-Crostic No. 36
Two ILec.e.nt -tltag.ic. a.c.cA..den-tt. __ ._
add up :to an evr.iec.o-incidenc.e
li
L
.
86-36
Dec 81 * CRYPTOLOS * Page 14
UNCLASSIFIED
UNCLASSIFIED

P.L.
I I
DOCID: 4011945
Dec 81 * * Page 15
UNCLASSIFIED
DOCID: 4011945
Fe!! 8FFleIlt:l:i ass 8t1l:i
P.L. 86-36
(U) I'm a Traffic Analyst. Several years
ago I began to work for the person I respect
most in that field and share in the develop-
ment of what has come to be called a Traffic
Analysis Workbench System. What ever comes of
that eff9rt, I'll always be grateful for being
included.
(U) Having resisted the urge to vent my
excitement on paper for this long, I thought I
had it under control. Actually, I have been
writing this piece all along. Part of my con-
trol mechanism was simply typing my thoughts
on the screen and then hitting the delete but-
ton. That may happen to this version and you
will be spared once again. I'll tell you what
"set me off" this time a bit later. First, let
me tell you what I've been so enamored with.
(FSYS) The idea is a relatively simple one:
In terms of technology, the TA field is, and
has always been, behind the power curve.
Regardless of what high powered machines
exist, and in spite of the fact that some
extremely sophisticated machine applications
have been designed for analytic purposes, the
analyst is still behind. There are several
reasons this situation exists. However, it. is
primarily because analysts are directly
by1 _
In pursuit 01:
~ ~
~ or the past two years I have been
fjJ
involved with a project that has
given me more, in terms of psychic
income and pure excitement, than
(U) perhaps any work-related activity
that I remember. When you feel this good
about something it seems natural to want to
tell everyone else and share the excitement.
Sort of a "look what I found" feeling. Of
course, when you feel excited about something,
it is difficult to know whether or not you
have something worth saying and can remain
objective about it. Nothing makes you feel
quite as foolish as discovering the wheel only
to find out that you were the only one who
didn't have one all along.
(U) I've followed -- from a distance -- the
articles, letters, and symposiums decrying the
diminishing number of analysts, the dilution
of the career field, and the increasing work
load. I really have nothing to add to the body
of literature that has grown around those
themes. I would like to note that some reason-
ably intelligent people have advanced them.
Conversely, some reasonably intelligent people
made the decisions that led to the described
conditions.
Dec 81 CRYPTOLOG Page 16
F8!! 8FPleltb ee! 8Nb
DOCID:
, ( I
4011945
I
S8NFIBSH'fUb
dependent upon their machine support person-
nel. A few have managed (mostly out of frus-
tration) to learn one computer system or
another and support themselves. The problem
with this ia, if they were any good at it,
they were usually lost from the field.
+eT It's time the analyst was given some
help. Not to catch up to technology, just to
keep from getting further behind. Given the
costing cycle, the procurement cycle, and the
installation cycle, I'm convinced that 'catch-
ing up is not possible. 121 possibleI The con-
cept of a TA workbench involves installing a
terminal on the analyst's desk. Read that
againl Not down the
hall in a "machine room", not in a corner of
the basement, and not around the corner where
it "won't bother anyone" \.' ...
ll.U.
.....__-' This put,s the analyst in a posi ti.on to
access the major data bases, where the/daily
traffic as well as the technical working aids,
reside (hide is a better word). With the
minals.2n they will have constant
access to their material and perhaps approach
the paperless environment.
(Feee) Under the umbrella called PINSETTER,
we have been proceeding along a development
path that will hopefully lead to/the /kind of
help the analyst needs. Because the most
precious computer resource is the programmer,
the analyst must be released from depending on
him for every minor need. This is true for
several reasons. First, the analyst needs to
be able to access his data, pr0gess that data,
and change those processes Without having to
wri te memos, generate specifications,
write justifications, wait for software and
then participate in debugging. Second, the
programmer, as a resource is too valuable to
be tied up with changing sort specifications
every time an analyst needs a different out-
put. Lastly, the plain facts are that we have
a terrible time retaining good programmers. No
sooner do we develop a good working relation-
ship with a programmer, and he'
begins to understand something about an
analyst's job, than along comes a better offer
and he's gone.
(U) From lil machine standpoint, meeting
these goals requires a system that is easy to
learn, fleXible, and provides a reasonable
response time. By "reasonable", I don't mean
instantaneous. Most analysts can live with an
execution time that is not measured in nano-
seconds. Most of our work has taken place on a
PDP-11/70 host using UNIX as the operating
system. UNIX is a high level language that was
developed by Bell Laboratories. It meets the
above criteria plus it is very forgiVing to a
klutz at the wheel.
We have found that most of the
processes that a Traffic Analyst needs to be
able to do can be accommodated with the UNIX
package. Where it was found lacking or ineffi-
cient, the solution has' been provided by a
unique working relationship with a small group
of hiJl:hlv talented DrOllrammers in The
(U) To digress for a moment, the reali-
zation that certain processes are simply
too big for TSS applications is impor-
tant to maintaining a proper perspec-
tive. This determination must' be made,
and large "number crunching" must be
performed where they are most effi-
ciently handled. However, the process
can often be executed where most effi-
cient, and the results passed to where
they can be best used, on the T55. I
might add, in two years of handling TA
. processing, it has been necessary to
"send out" only one job for actual exe-
cution on a "big" machine. Of course,
many of our extracts from major data
bases are "preprocessed", prior to
transfer, to make them more TSS
friendly. But I discount this, since it
is largely "invisible" to the requestor.
If the solution were apt to prOVide
a useful "UNIX-extension" , we woul,d request
T333 help. The results have been the most
rewarding part of this experience: generalized
UNIX-like utilities that solve analytic pro-
cessing problems. The big plus? who
EO 1.4. (c)
P.L. 86-36
Dec 81 CRYPTOLOG Page 17

i
l:i
iii
.
,.1[1.: .: J'I
.II:
DOCID: 4011945
P.L. 86-36
EO 1.4. (c)
knows a little UNIX can use them. On the
other hand, if the solution appeared to be
problem specific, we would attack the problem
with our own resources. The results of these
efforts have proven equally rewarding. Based
on our own experience and some operational
testing in analytic elements of A3, B2, B5,
G6, and G9i I'm not sure if a more effective
analytic tool than UNIX could have been
designed if that had been Bell's intent. This
leads to a philosophical difference in user
support design.
There 1s a mask-and-menu school of
thought that holds to the belief that the user
should be led through the processing cyale by
the software. A.menu is presented with a few
options to select from and a mask provided
through which to make alterations. These H&H's
believe it is best to protect the user from
the complexities of the system and protect the
system from the klutz at the wheel. It has a
place. I would look to this area for the type
of handling necessary for, perhaps, TEXTA
updates.
(U) Another approach is to provide the user
with the modules necessary to manipulate the
data, a high level language to package the
modules, and the ability to communicate with
other users and peripherals such as high qual-
ity printers. Basically, a sort of Procedural
Applications Language that is not unique. to
Traffic Analysis. Perhaps a Universal Pro-
cedural Applications Language approach. The
user is free to design personal processes and,
more importantly, those processes at
will. Users are not dependent upon the pro-
grammer for every minor modification, routines
do not have to be recompiled after each
change, and the results of the changes are
immediate. I believe UNIX meets this chal-
lenge.
(U) The H&H approach "keeps the analyst (or
user) dependent upon the programmer for modif-
ications. Thus, preserving the problem of too
much demand being placed on a resource that is
already over taxed. The solution to the demand
for software packages has all too often been
the letting of contracts, at considerable
expense, to develop processes that a few
analysts, skilled in a handler like UNIX,
might be able to get along without.
(U) If our own resources were concentrated
in a manner conducive to the development of
generalized handlers (a Universal Procedural
Applications Language), and perhaps a bit of
that contract money concentrated into reward-
ing the good programmers we have left, we
might be able to come up with better analysts
and better programmers. As a by-product, we
might be able to handle the workload with the
number of analysts we have and do a job
of it.
(U) So, what was it that set me off this
time? A few days ago, while a
few system capabilities to a potential I
was walking through the steps of a UNIX shell .
file (merely a collection of UNIX commands
that eKecute sequentially and perform some
proceaa) and he asked me if I "wrote this pro-
gram". The words startled me. Wrote a "pro-
gram"? Me? I'm a Traffic Analyst, I can't
"program". My rather bumbling answer was some-
thing to the effect that this is really not a
program just a collection of instruction$to
perform a certain process on this computer.
After he left I put the shell on the screen
and read it a few times. By gosh, a few years
ago I would have called that mess a program
myself. It "looks" like a program. It "act-$"
like a program. And, my extemporaneous answer
wasn't too bad a definition of a program.
(6 I had to pause and reflect a bit. I
put that shell tOllether in about five minutes.
what does it do?1 ....
Based on past experience in trying get a
process to do a select of this nature, and
going through the "channels" to get it; this
"quickie" shell seems fairly. powerful.
(U) I think I've found a faster horse, I
probably couldn't keep up with younger women
anyway, and I'd rather have a cold beer than
older whiskey, so if anyone knows someone
looking for a "programmer", I'll settle for
two of four.
Dec 81 CRYPTOLOG Page 18
II/dfBbB 'lIA eSUill'f SIIANIIBbB SlIbl
OCIO: 4011945
I
86-36
Whatisit?
Where is it going?
byl Ipl
TEXTA
means Technical Extracts fran
Traffic Analysis, and represents
an agreement between four national
centers concerning -
- (!I)-
the exchange of basic traffic analytic
informat ion1
the sharing of a cooroon, uniform record-
ing and labeling system of traffic
analysis information about CCJoIINT targets
worldwide1
a catlllOn book of rules, the 'lEXTA manual,
which the four centers accept as the
authoritative description of how the
TExrA system operates1
the highest, mst accurate level of
knowledge on a target carrnunication.'
EO 1.4. (c)
P.L. 86-36
Dec 81 * CRYPTOLOG * Page 19
II*NBIdl ,'tiL B8A1RW' BIIAI'II'IIWl 8NU
DOCID: 4011945
How Current Should TEX'l'A Be? (U)
Using TEXTA for Collection Steerage (U)
.L. 86-36
1. 4. (c)
Dec 81 * CRYPTOLOG * Page 20
ttttKI't! '11\ eetlHl'f eth'dftfEI:iS elfl:i'l
1. .L __
DOCID:4011945
Dec 81 * CRYPTOLOG * Page 21
EO 1.4. (c)
P.L. 86-36
HMiBhB \':l:A eaU:l:tl'f e ~ ! i B h S a!ih
DOCI D:
Where is TEXTA Going? (U)
0 _
(U) such "low validity data, eyen/ though
current, should not cause the 9lder, but more
authoritative, analytic resl,llts to be erased
fran TEXTA - they should/coexist side by side
in the record. '!heeaae with which data can be
lost through erasing and/or spillage./ is a
major problem in coop1ter based data systems
and il:$/One that will need special attention
for'I.'EXTA in the future. '!he future system
should be set 1,1p to insure that/no data is
erased unless a back1,1p record (suGh as micro-
fiche) is first generated. Data no longer
wanted in the current file, because it is
either out of date or superseded, should be
shifted either ,to "near storage" or "far
storage" , depeooing upon ./the likelihood of
having to retrieve it at/ sane later
Near storage is defined as sane machine-
retrievable form, such as tape, where the data
can be retrieved relatively easily, whereas
far storage is defined as sane form that is
essentially not machine-retrievable, such as
microfiche, where .the data can be retrieved
only with great cost and difficulty. (The
cost of "repoking" or otherwise retrieving
data fran microfiche may make such efforts
rare, but it.would be unwise to rule them out
sanetimes the thing MUST be done,
even if it has to be done entirely by hand.)
Data might typically move fran current on-line
storage/to near storage, and then after a
specified period of time, to far storage.
(U) Future 'I'EXTA will need an integrated
of "audit trails to accOlllIlOdate the
variety of levels of data that will be in the
system. At the minimum, the system will
need -
the date of the action which changes,
adds, or deletes the infocnation.
the source of the action (at a minimum,
the organization submitting the action,
although at some locations the initials
of the analyst might be needed).
the validity of the information involved
in the action.
P.L. 86-36
EO 1. 4 . (c)
Dec 81 * CRYPTOLOG * Page 22
- L __
DOCID: 4011945
I
Dec 81 * CRYPTOLOG * Page 23
EO 1.4. (c)
P.L. 86-36
\\
l W I B E ~ E 'lIA eStHtI'f elhldltlEbB Stlb'[
DOCID: 4011945
the recipient).
I
P.L
EO
Should the Future TEXTA System
Look Like? (U)
')${fA?
chen sie . <
S'pr
e
C? .... c

0
c'"
?@

,,"-

<;-.0
Ha
b1a
TEXTA7

:fo''O
@J'Q
<;0
!JiI,.
'il

Dec 81 * CRYFrOLOG * Page 24
. 86-36
1. 4. (c)
UMIBbB HA eetfftl'l' ell:AtlUBbS BilbY
DOCID: 4011945
Wlth a canputer canplex. No user need be kept
out of the system, or reduced to a second-
'P.L. 86-36
class citizenshipiIlt:.QesystembeCause he EO 1 4 (c)
... \ . .

(U) The system shoUld provide for some kind

of audit trailing,so that a selected class of
user (i.e. sane but' not all) can determine:
* who put a,particular piece of data into
the system?
* when?
... what "validity" was ascribed to it?
"low validity" entries cannot
erase/replace "high validity" items
already
several variants (different sources,
different validities) may have to
coexist in the system for " extended
periods of time.
* items erased/replaced (are not typically
thrown away)
are retained in the "dark end" of the
record (or in "near" storage) for an
extended period, then are stored in
microform (or "far" storage) and
remain available for audit trail pur-
poses.
Dec 81 * CRYPTOLOG * Page 25
4011945
onLY
PLATFORM:
How Did You Say
That Works?
P.L. 86-36

users to sit at their favorite ter-
minals, and travel allover the
(U) world to accomplish their assigned
task. All that is required of them is a few
simple standard commands. Oh, wouldn't that
be great I
(F6ij6) Wait, aren't you talking about the
PLATFORM network? Isn't that the way it
works? Almost. We are close, but we aren't
there yet. Simple things, little things, that
appear trivial when looked at individually,
work together to cause most of the current
PLATFORM user frustrations. The general idea
of PLATFORM is a good one, but it seems that,
for those who actually have to use PLATFORM,
something has been lost somewhere:

Terminal characteristics that differ from

one terminal to another as well as from
one host to another.

Terminal functions that can be 'Used on

some network hosts but not on others.
Network capabilities that are supported on
some hosts and unavailable on others.
Response times on network-connected termi-
nals that exceed normal expected overhead.
Limitations on the number of connections
that hosts will accept from the network.
Multiple logins between hosts and again
for processes or applications once logged
into the hosts.
that are
host to
protocols
from one
Call-ups of network
slightly different
another.

Dec 81 * CRYPTOLOG * Page 26

DOCID: 4011945
FeR eFFlel!b 68! 6NbI
P.
EO
\""
(U) Let's make it a tool that serves those
charged with carrying out the mission of this
agency.
Dec 81 * CRYPTOLOG * Page 27
L. 86-36
1. 4. (c)
DOCID: 4011945
POft OPPIeIAL ONLY
In The Name of Efficiency(u)
review by
1 lpI3
(U) BOOK REVIEW:
Joan M. Greenbaum, .In.tbl. li.iI:rm.2! Efficiengy,
Philadelphia: Temple University Press, 1979
(NSA Main Library, QA78/G82)
(U) The subtitle of this very interest-
ing and provocative book is "Management Theory
and Shopfloor Practice in Data-Processing
Work". It presents a frank and (to my eye)
refreshingly critical and challenging review
of the history and social context of computer
programming and operating. In the view of the
author, this history has been marked by a
tug-of-war between management and the data
processing workers. On the one hand, manage-
ment has been uncomfortable with the freedom
which programmers had in the early days of
computers, and has found ways to limit that
freedom, with increasing success. Program-
mers, on the other hand, have fought back to
preserve the work satisfaction and status of
their occupation.
(U) Here are a few brief quoted passages
from the book, to illustrate the approach.
Ms. Greenbaum begins her Introduction with the
follOWing personal scene-setting: "Back in the
1960s I was a computer programmer. Like most
of the 200,000 or so other programmers, I
enjoyed the work - particularly its opportuni-
ties for diversity and challenge. Compara-
tively high-paying, computer programming
offered high status because its skills were
little understood and in great demand. By the
early 1970s some of the craftlike characteris-
tics of this work had begun to change. The
changes, like most day-to-day happenings,
appeared quite slowly. But as they began to
increase in tempo, it gradually became
apparent that work activities once controlled
by data-processing workers were no longer in
their control." [po 3] She continues, "In what
began as a personal stUdy, I set out to
explore wbat was taking place in data-
processing workshops and why it was happening.
Many have said that the changes in the work
process were just the results of 'normal'
86-36
changes that occupations go through as they
mature What was most noticeable about
the changes in each occupation [so affected]
was that they were anything but 'natural';
workers fought against these fonas of change,
and managers had a hard time implementing them
The reasons for changes in the work-
place are not always the reasons that appear
on the surface." [p.5] She states that her
purpose is not merely to "bemoan the lost days
of craftlike activity," but to reveal the
underlying reasons behind the changes and
enable workers to understand, influence, and
regain control over the workplace.
(peye) I was particularly interested by her
review of changes in the field; I myself
remember many of them as they happened here at
MS!. What was once a single profession, "pro-
gramming," (where a "programmer" carried out
all phases of the task from problem definition
to operational running of his program) was
fragmented into disciplines performed by very
different sets of people: operators, program-
mers, systems analysts, and keypunchers. The
separate disciplines were often divided from
each other by distrust and hostility as well
as by the physical and organizational "walls"
of the "closed shop" philosophy which was
popular with management for a while. The
advent of operating systems removed much of
their new power from the hands of one of these
new groups the computer operators. All
these changes, Ms. Greenbaum convincingly
maintains, were results of deliberate efforts
by management to "divide and conquer" the
recalcitrant data-processing workforce and
"rationalize" their work, in order to bring
them under management control. Interestingly,
she makes a clear case for the origin of the
"professionalization" of programmers in a
management initiative, and claims that this,
too, far from being a desire of programmers
themselves, was a step toward control of data
processing workers by management. Structured
programming is another obvious landmark in
management's strenuous (and all-too-
successful) efforts to remove inconvenient
degrees of freedom from the programmer.
Dec 81 CRYPTOLOG Page 28
peR epPieitb as! 6NLY
DOCID: 4011945
FQR 9FFiSihh BaS 8NhY
(F8e8) Most of all, .the IBM 360 systems
made a dramatic break with the past. Ms.
Greenbaum says the following: "Those 'of us in
the field at the time of the introduction of
the System 360 tend to remember it well, for
almost overnight a firm division of labor
occurred, not by chance, as it seemed to us
then, but by clear design. Although computer
work had been divided by task in the 1950s,
many activities had overlapped a good deal.
In particular, computer programmers and opera-
tors would meet in the computer room, which,
like a social hall, offered the opportunity to
exchange techniques and ideas. The installa-
tion of the System 360 provided management
with reasons to change this. One of the first
rulings to be enforced was a prohibition
against programmers entering the computer
room, thus isolating the two categories of
labor and cutting off exchange of functions
and rigidifying job olassifications." Ms.
Greenbaum analyzes the theory whose applica-
tion by management brought about these changes
in Chapter 3 of her book. She provides provo-
cative treatments of "Shopfloor Practice,"
"Labor Process," and "Worker Behavior" - the
interplay of workers' responses and resis-
tances against management's initiatives in the
daily operation of computer shops. She covers
a great deal of very illuminating material
concerning the workers' perception of their
jobs and the ohanges enforced from above by
management. I found many vivid echoes in my
own memory of these perceptions as I experi-
enced them in our own NSA computer installa-
tions, since I began in the "craft-like" times
of ATLAS I, and lived through the changes
accompanying the 704, 7094, and Systems
360/370.
(U) The author leaves us with an unex-
pectedly hopeful conclusion at the close of
her book. She gives much weight to the
efforts of data processing personnel in
creatively remaking their work situation, and
in partiCUlar finding new ways to cooperate
and communicate with their co-workers and thus
reclaim control over their work activity and
restore challenge and satisfaction to their
jobs. She concludes that "data processing
workers have developed workplace activities
and cooperative work practices that stand in
sharp contrast to the rationalized bureau-
cratic hierarchy imposed by management. We
are told that human nature is competitive and
indiVidualistic, but data-processing shopfloor
actions contradict this. Effective data-
processing work is usually accomplished by
workers who help one anothep by sharing
knowledge, skills, and tasks. By sharing
knowledge data-processing workers have
created, in effect, their own shopfloor cul-
ture that gives workers at least the ability
to tolerate the contradictions they face every
day on the job . . When I first began this
study I exam.1,ned management justifications for
efficiency and tried to compare these to what
was actually taking place in the work environ-
ment. The more I looked the greater I found
the differences between management and worker
strategies for workplace activity . . Work
does n21 have to be organized to control human
behavior. Efficient work activities can take
place without the management ideology of
social control. Examining workplace activi-
ties begins to point us in the direction of
understanding other forms' of work organiza-
tion."
The author's high of
cooperativeness and creativeness among pro-
grammers agrees well with my own experience
when I was a 'full-time applications program-
mer. We shared ideas, helped each other to
"debug" programs, shared labor (piaking up
runs), etc., and found ways to forgather in
areas near counters, key-punches, etc. to
exchange news, techniques, tips, and aid. We
often had to do these things in spite of
management's frowning upon our apparently
unstructured activities, and our uses of
spaces and facilities intended by management
for other purposes.
seems evident that these con-
soon become more crucial than
ever. It will very soon be possible for mUCh,
if not all, programming and computer-aided
problem solving to be done via remote termi-
nals. Soon it will no longer be in any sense
a practical or physical necessity for data
processing workers or computer users to be
located all together in one building. The
only thing that might continue to force vast
numbers of computer workers to be herded
together in offices from nine to five, five
days a week, would be the fear of management
that any other arrangement might result in
their loss of social and behavioral control
over employees. I recommend Ms. Greenbaum's
challenging book as a starting point for
thinking about some of these issues.
Originally published in the October 1979
Newsletter of the Special Interest Group
on Human Factors, Computer and Information
Sciences Institute.
Pi-Jan 82-S3-12558
Dec 81 CRYPTOLOG Page 29
FeR eFFiSihb yas 8Mb
DOCIDcct4iii 945
....-.
----=-=----- ..
-- - 11 CJ'----_'---_""-
'I'IIIS BOSf:JMBN'I' SON'I'AINS SOBB\\TORB
- --- ----