THE INTELLOFAX SYSTEM (THE CIA LIBRARY AND THE MACHINE DIVISION)

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP84-00951R000300040008-9
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 54
• Document Creation Date: December 9, 2016
• Document Release Date: December 26, 2000
• Sequence Number: 8
• Content Type: REPORT

Approved For Release 2001/08/01 : CIA-IMNfoligKgR000300040008-9
R-
A
T # 1 THE It7TELLOFAX SYSTEM
(The CIA Library and the Machine Division)
I. EARLY DEVELOPMENTAL HISTORY (1947-54)
A. Cbjectives and Ect1.14
In providing a central reference service to CIA and
the intelligence community, the early managers of the
Agency recognized the need to develop a machine capability
for indexing and retrieving a staggering quantity of intelliu
gense documents. The resulting Intellofax System, which
evolved jointayby the Machine Division and the Library,
was unique-. no other goverment agency, no university or
library and no commercial firm had anything of its type
in operation. The name was coined by Dr. Andres in 1949
to describe the system which combined IBM and facsimile
reproduction technique d for intelligence documentation
purposes. Later in common parlance, the word was used
not only as a noun-L' (the Intellofax System and the Intello-
fax files) but also as a verb form (intellofaxed and intello.
faxing for the indexing aspects) and became a household word
in the intelligence community'
The actual authority for establishing the Intellofax
System appeared in an ORE Instruction # 3.1-47, entitled
25X1A2a Functions of the Reference Center dated 15 July 1947o
111111111111111 Assistant Director of ORE, charged the
Central Tadao (later the Machine Division) and the Intelligence
Documents Division (later the CIA Library) be
Approved For Release 2001/08/01 : CIAleireier1R00030004000
GROUP
Excluded from automatic
downgrading and
declassification
DITELLorthapjmil 5orRelease 2001/08/01 : CIA-RDP84-00951R000300040008-9
DEGRET
25X1A5a1
25X1A5a1
After numerous meetings withnnd investigation
of other companied, such as
the machine
25X1A5a1 experts opted for IMInd a contract was let in January 19480
25X1A5a1
BY July Illigad produced the first of the Library Recorders
and had completed the final design for the IBM card scannee.
Both awaited OCD approval? Experimenting and testing continued
and in January 1949 reported favorably on the equipment,
commenting that
25X1A9a
25X1A1a
it was indeed gratifying and thrilling to
see the first phase of this development
actually operating and with such fine quality
results. ? it illustrates th501715pt
effort that the people of the MJ..11 have
been and are putting into the job.
Progress reports were prepared periodically throughout
the first six months of 1949; test runs were made during Juno
and the equipment was finally accepted in July. The Projecta
Review Committee on 27 July 2949 approved an amendment to the
original contract, which had been in the amount of
to the amount of $
25X1A1a
The Intellofax Card, or Faxcard, was an IBM punch
card of standard share and dimensions which bore on its face
up to 200 words of printed information, the so-called bibliographic
data: source, country, date, title, possible abstract, pagination
and security classification. The corresponding coded and punched
data appeared at one end of the card. The cards were sorted, selected and
Approved For Release 2001/08/01 : CIA-RDP84-00951R00030004
RO@ON*W.
III U
?
u ed tromPalutornalic
downgrading and
declassification
INTELECTAX--page Lt.
Approved For Release 2001/08/01 : Cfeltifti4-O0i9i5lR000300040008-9
arranged by standard IBM machines; and the printed inf.'orma-
tion on the s elected cards was transmitted and reproduced
by facsimile process?
The equipment delivered in May 1950 was the second
prototype resulting from the developmental engineering
begun in January l9480"Shakeown't tests were still. being
conducted in mid-1951 concurrent with actual usage0
25X1A9a
an Office of Communications employee 25X1A5a1
(and foxner an engineer wIt was on temporary
duty with COD and placed in charge of the Faxcard equipment.
He wrote to (chief of the Machine Methods Division since
September 1950) that since the equipment was not standard
equipment, additional development was anticipated before
the stability of the equipment could be placed in a class
with that afforded by isting teletype machines.
At the same time that test runs were being made
on the I EM5 tipment JEN nvestiated the potential
25X1A9a
25X1A9a
25X1A5a1
25X1A5a1
25-idA5a1
use and availability of therprinters which would repro-
duce/Printed, typed or written data by a heat' procese0
nese aras responsive to OCDts urgent need
for this type. of equipment and agreed to build and demonstrate
a prot'Abeeof the machine by Ju37 1949? This was the basis
for the/
firat Intellofax tapes printed continuously onto
thermofax paper, somewhat similar to though smaller than,
CROUP 1
Excluded horn automatic
Stiel,RIEF
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
INTELLOFAX?page 2
Approved For Release 2001/08/01 :4SILIgagig0951R000300040008-9
25X1A9a
25X1A5a1
(1) index, by business machines procedures, the
subject ratter of all available r rts, and
other document4 of a foreign v elligence nature
and (2) 4assifiy and cat gue all intelligence
25X1A9alrateria2.s of a foreign intelligence nature to CID.
Chief of Central Index, was given the responsibility
for organizing and developing the initial essential steps toward
oetablishing a central indexing and filing system, in conformity
with an earlier ICAPS recommendation in March 1947. It soon
became apparent that no existing eqnipment would be capable of
meeting the needs envisaged, Although an IBM punch card offered
great flexibility and speed in the handling of thousands of cards,
each of which would represent a particular document, no card
would carry enough printed data to supply the researcher with
IY
titles and descriptions of documents. During 1947 111111 and 25X1A2a
25X1A5a1
his deputy, met with top management of
to discuss the possibilities of the us* of
25X1A5a1 standard machines and the adoption of these machines25X1A5a1
to the documentation problea. A said that
his ccmpany would be willing to cooperate with IBM in adapting the
Telefax machine to automatically reproduce bibliographic and subject
abstract data typed on IBM cards onto any type of paper including
a duplicating mediemoThis would answer the problem of preparing
accession lista and lists of abstracts requested, (Management
originally planned for a daily accession list of those intelligence
documents receivr5d and indexed, all of which would be abstracted.)
....?1141:,, enter ORE to
Approved FoirfeltaAc_et9#161/1 if* wi, l? ?51
7 Julysi su.: Es ;.- ,n5 ? ant*
U. (in Chrono 1946-47 50-98/8)
Central Index
g Procedures.
a1pfo0i444ki
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
25X1A6a
The Intellefwc tape; as it was known throrgh the entire
Intellofax history, was originally a 4 inch wide tare prepared
the facsimile process* The Litellofax punched card was fed into
which optically scanned the
a transmitter the printed informat.,ione
A ree iver received signals from the transmitter; the printed inforilation
was impregnated into a chel.Tically treated tape which was dried hy
heat pr oc s s The-- resdAting-eontinuourr role- 'of deSid171pe
144-.4d-andr-ultimpts1,7--giverr-to-nirfel:CCeStelT
The early OCL managers had hoped to o lectronically
tran.srait the Intenof.ax information to requestem In their o,,,rn
_office locations, As of lf; 1.I.ay 19502 6 transmitters and 12 receivers
had been del iver ed. Expericntation cont inue dtt hro 'all out the s ujiner
months and the first bransmission was strictly local transmitter and
receiver side by side ia the Y`achine. Division* One receiver as placed
in the Branch Library, but security considerations and
technical problems of transmission were responsible for not
continuinc with what seemed like a .Utopian transmission phasqle
The production of Intellofax tapes in the 1950to and -
1960's first by facsimile process and later by the Oard List
Camera and Photostat Expeditor always remained in the Me chi
oldet /
Division* The completed taped lox_ MO.X later in booklet form
were delivered to the regnester hat, Via electronic trano;ission
but h7 hand*
* Photo of Transmitter and Receiver
GROUP 1
Excluded from automatic
downgrading and
dectossification
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
TNTEULF/0,--page 6 "AliSlikurgrnm.
"36 Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
Bp Codilr Schemes
armot.*
J. c The Intelligence Subject Code (ISC)
In conformity with ICUS wishes steps were also
taken by the Central Index- to prepare a unified subject
25X1A2a
classification scheme0 rote to in 25X1A2a
July 1947
Although the Reference Branch has taken the
initial steps in the direction of establishing
central indexing and filing procedures, any
unified acceptance of the end product of these
investigations will depend upon joint action
of IAB and cm representatives and the agaaciest
final acceptance of the sysbam decided upon
25X1A2a
On lit. July 1947
entered on
duty as Chief of the Classification Unit of the Intelligence
Documents Division towork with the Central Index in
developing a classification schedule for CIGe
It was soon evident that the Nr Departrentts
Basic Intelligence Directive (3ID) devised during World
War II for collection purposes (although it had been u.d
ZAIA6a
for classification of documents in the C-2 Library in
immedlately after the 1:7,-tr) was not adequate. The subjects listed
in the BID were nat sufficiently comprehensive to cover the
wide range of subjects in intelligence documrts because it
had been devised for Arriy purposes ally0 :The economic, political
CECRIA'f
GROUP I
Eluded from automat-1
downgrading and
daclamificalion
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
ICIErI&O--page 7
"kakpproved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
25X1A9a
25X1A9a
25X1A9a
25X1A9a
and scientific sections were woefully weak c It was decided
to prepare a list of' subjects which wonld include those
con6eined ia the BID, the Navy Yonograph Guide, ths-
abridged Dewey Decimal system used by the State Depertment,
and for scientific subjects, the Voge Classification, prepared
and used by the Joint Research end Intelligence Board (JRBD).
made visits to the parent organizations
using these classification schemes.
By August 1947 iassification Unit of 8
people had completed a general frarework of an all?inclusive
classification schedule with the assistance of 25X1A9a
111111 a classification specialist from JRDD, (OCD tried un?
successfully to recruit as a permanent employee). The
major subject categories included: Army, Navy, Air, Political,
Economic, Sociological, ScLentificm Geographic, and Biographic.
On 22 August a fmniliariation meeting was held with duay
appointed representatives of the three servies, The participating
IAB agencies agreed to develop and/or reviEe their respective
military categories in the DID. To those entegories would be
added the GIG contribution consisting of the nonnilitary
subjects. Because the War Department was not inclined to
change the numbering system of the BO D (8 digits), it, ae to be
used as tile nucleus of 'Ole nut, classification syAen.
GROUP 1
Excluded from automatic
downgrading and
declassification
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
1SC
CECILILT
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
25X1A9a
was not very enthudiastic about t he coopertfLon from th..7:
25X1A9a
other agencies. -e and
had visited the State Eepartrent
Libra:i-ans who ve3col,,Jd a comprehensive expansion of the Argv, 71-vy
and Air subject classification, but felt that this expansion should
be incorporated into the abridged De7ley. The repro3entatives of the
1A8 agencies seemed to feel that what CI3 was trying to do with a
new classification would replace the classification which each agency'
was usingo Thi:-; was, of courses the ultianbv5 alms but it would not
be realized mien partiallj un4-11 the Af.- force adopted the intelligence
Sub'ect Code in 19540 iLach representative took a cosmic view of the
fields which were of primary interest to h:ts agency and argued that
the whole structure of intelligence would be imperilled, by any
2/14
deviations with its own scheme. ?
So the Library set about continuing with its awn classification
The first edition of the Intelligence Subject Code (henceforth
referred to as the ISC) was dated 15 March 1948. The Preface indicated
that the edition was provisional and that the subject headings ware
intentionally kept rather general so that expansions and revisions
could be made as experience required. There was no index to this
first edition. A biographic or "Vhois Ilho? clgss which was in the
original outliae was deliberately omitted because of the Biographic
Intelligence Register of the hoference Center. The main classes ?,11(1
the number of notations (codes) were:
000 internaticnal Situation (32)
100 National Affairs (120)
GROUP I
Excluded from automatic
downgrading ond
declassification
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
INTELLOI.?1\X?page 9
,-
- Approved For Release 2001/08/0141111A1000300040008-9
200 Aznr (139)
300 Nay (181)
400 Air Force (83)
500 Weapon: a nd Scientific Warfare (44)
600 Science and T chnolog-y (82)
700 Geography and Economics (232)
800 Social and Cultural Forces (67)
Total notations: 980
Each of the 0) t categories was broken down to provide:
as nearly as possible: for the needs of the agency rThisfly
concerned-- the A my: Navy and Air iections follaving closely
the patterns developed by the three serve es for their con uce. The
other sections had been worked over in detail with the ORE
units chiefly concerned+)
Chapters 100 through 800 retained their overall
subject outline until the complete revision of the ISO in
November 1960Q Further chapter sub-divisions appeared throughout
1948: but it was not until November 19h8 that the 600 and 700
sections were expanded to the full six-digit capacity -ilottd op the IBH
card A relative index (alphabetic31) %.n PC printed a the same tine0
440?101r1 T
GROUP 1
Excluded from automatic
downgrading and
declo.ification
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
INfipi,L0fAX-i. page 9a
IC Approved For Release 2001/08/01 :armilig0410951R000300040008-9
25X1A9a
3-0B took
in the Library .
who had reported for dirtyA on 9 February
over from in mid-1948 as Chief of the 25X1A9a
Analysis Section (formerly Classification Unit),
(She---,?,emainod head of the input or classification effort for
the Interlof=. System almost a) years until the demise of the
system the end. of 1947.) She worked closely with analysts of OW
ORR and 031 in the continuous revision process during the first five
to
years, ensure more effective organi ation of the information in
documents. These research analysts pointed out deficiencies in
certain subject fields and suggested appropriate changes. Most
suggestions benefited and improved the ISO; others refelcted only
parochial needs of insistent and narrow-in-outlook requesters wh,
raised their subject specialty out of all proportion to the entire,
scheme of knorledge.
r The latter type of
requester made one section of the ISC look ridiculous: the
subject code :for Plant Pathology (632.4) was sub-divided into
68 cufferertt codes f or wheat, rye, barley, oat nald miscellaneous
crop diseases with the names in English followed by the scientific
term in Latin.
The 1949 ISO resembled the original 19)5 edition
only in the 8 major chapter headings. Within each chapter
much restructuring took place. A new heading or Communism w as added
and th, 114 section became the most widely used and remembered throughout
the
the book0 geography was moved from 700 to the 600 chapter. In 1950
Approved For Release 2001/08/01 : Ce4F4111406951R000300040
GROUP 1
Excluded from outontolie
downgrading and
d ossification
INTI.TCAX--page 10
:13a Approved For Release 2001/08/01 M17p0951R000300040008-9
at the time the Library decided to catalog boohn
according to the ISO, a 900 chapter (Organization of
Inforintion) was added,
The history of the ISC from 1948 through 1967
was e hi3bory of change and hoped-for improveaent. 980
codes grew to 5,000, A review of the master copies of the
ISC during these years reveals many pages of revisions*
1..e45?wI-4iew editions were puolished in 195113 1957, 1960,
1962, 1964 and 1967. Changes in subject codes necessitated
Vox, preparation of new cards. The printed information
was transferred from the old card to the new card by means
of a heat process, -whereas the punched data was converted by
machine under punches to the new cofts. This was a the
process and caused backlogs in the achine
Division, (The biggest change to an entirely new 130
in 1960 did not involve conversion; threafter? the
Intellofax cards were kept in separate files--"A" file
from 1960 on and "B" Me-pre-1960),
All classification schemes have limitations) andthe
ISO was no exception? particularly since code expansion
was tied int,o the Allotted spaces on the IBM card.
pz,rsonne.17-eluays verked-cloJely-with thsa-Nachino!-DIvig'ion-rerM5nTel:
before-anything-ttniquewa5-adopbad. As mentioned earlier, the
fallAigit expansion of the 7f,0 chapter went into effect in
November 1948. By 1950 it -come evident that certain aspects
lEot ;MP 1 '? I
xtludad from outomotic
downroding end
d lassification
Approved For Release 2001/08/01 : - - 951R000300040
INTELLOFAYA
13-1AQd:/i-or Release 2001/08/01 Af@lfigagliga00951R000300040008-9
13C
of information could be uniformly applied to almost all
and equipment
commodity subject codes in that chapter. These tiactions" Were
production data; imports-exports; maintenance* repair*
replacement and construction; ,procurement* etc. The Library
and the Machine Division personnel worked out a unique
schime for affixing a modifier before the subject code.
A list of on to twc, digit "action" or prefix modifier
codes was established. The classifier entered them on the
code sheet by placing a slash between the modifier and
44-subject code. For example* the production of coal was
written as 11/73501. The slash appeared on. the IBM card as
an crir-rpunch in columns 1-6 (subject field)0
ilthis important change in the coding process
extended
eventually 1;5,71954 to svmst_ other chapters of the ISO.
Prefix modiers were applied to the military chapterszot
for such aspects as security* vulnerability* sabotage,
.order of battle* dpecifications and description of
railtary equipment.
Other coding devices were ilmaugurated. One
of the subject codes- 2.15 (Insurgent Groups) had no farther
breakdowns. At the request of the ORE/Greek desk at the
0--
tirne of guerrilla uprising in Greece the following
instruction appeared in the 1949 ISO: The 115 code may be combined
with the first cagitis of any classification number throughout
the ISO* e.g0 monetary system ofGreek
The same rule xduded from automatic
GOOUP 1
the ro ehare7,=.1Lnd
Approved For Release 2001/08/01 : C A 11951R0 030004 CIII8-u
-INTULCaia - -page 12
a4,400LmilTem
'SC
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
-Codinvg specificity was alsoladlievedin another manner,
was
s. A list of languages, minorities and cultures
prepared and coordinated with Andrews (his specialty)
and the 3 digit identiricqbioa could be combined
with either the 117 code for minorities or the 876 code for
foreign languages For exainple, the, Italian minority was
coded 117.119 and the Italian 1.-mgru 'vas coded 876.119.
Statistiie compiled for the Inbellefax System
indicated that an average of subject codes were assigned
per document.
"SILigElal?romTm.?
1-6-iaTir;-T---
Excluded from automatic
downgrading and
declassification
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
UlTELLOYAX?page 13
Are Aptiroved For Release 2001/08/01 : ar-M0951R000300040008-9
25X1A9a
2. Area Classification'
1,111111Ichose the Ari Map Service (AMS)
Library Classification as the best at most adaptable
system for ceding geographic areas. According to thici
system the world was divided into 26 main divisions, A
through Z. Each alphabetic division was further
subdivided, moving from r:l.ght to ]eft with a numeric
designation. For ercample:
M Europe
7:1M Scandinavia
11M Denmark
21M Finland
-01M Norway
4:114 Sweden
1)41M Northern Sweden
241/4 Southern Sweden
MIS did not maintain its area classification on an
up-to-date basis; therefore, the Analysis Branch (the Section
became a Branch a arly in 1950 was constantly e Xpanding the
code and updating it to Epecific Intellofax needs. When
India was divided into India and Pakistan in 1948, the f ormer
code of 5U became EU for India and NU for Pakistan, with
farther subdivisions for both countries?
geographic concepts and some typos of country relationships
were designated by Tre ans of a slath(i) which always followed
an area code. For exarrrpie, "A denoted a League, Confederation,
Axis or International Organization. Thus? the Arab League
was coded 6IciA; the United Nations became VA. (1 had been
GROUP 1
Wuded frotila,OwnoRc
dwwnmalland
dedaWScoflorl
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
IIAELLCVAX--page 114
:Area Codes
? ? Approved For Release 2001/08/01 :tralefie4/444.61R000300040008-9
established as the code for the World). /C de-
noted Commuhist influenced or dominated countries
and w as Vis ad effectively with the Eastern European
or Far Eastern blocs. By this device the Machine Division
could easily retrieve information on all Communist
44"
countries (other thanPUSER which had its own area code of N).
It was easy to segregate the Russian Zone of uermany (4M/C)
from West Germany (ItIVD)
Related Areas
The orlgja-,1 design of the 13M card P1 1 we d
the
foy;,,,k gits (columns 7-10) of the AP3 coete. Soon thereafter,
column 11 was allotted to the clash. Two years of experience
pointed up the inability of being able to show any area
relationships . This earn,: to a head with the 1950 Korean
Nero when it seemed necessary to be able to show sone
combination of Communist China,
446.?
Korea or the USA1 The entire punching area of the IEM
card (other than tho subject field which always remained
first, digits) was revamped, eliminating certain codes
did not seem nocessarfo such as dcty of publication and
date of information; and adding other codes, such as
digit abbreviated area codes to be used only as related areas
in columns 1.-220 The revamped card of February 1950 provided
USSR, North Korea, South
the
panIXT
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
[m GROUP 1
Kcludvd from automatic
downgrading and
TFLLOIAApprove15d For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
1NJX--pago
Area CO,c-,!8
'51f,CRET
space for two related areas of digits each. Area codes
contained more than ligits became abbreviated, such as
228M to SI for Spain. The classifier indicted them on
the code sheet with a parenthesis to alert the key punchers.
Example: N (01) (JM)? some relationship between
the-133SP. (N) and France (6M) and Russian Zone of
Germany (J1TL)
Area File
The advent of the Korean War also brought out the
neA for a separate file arranged by area. Requests which were.
coming in for everything on Korea could not bt-: answered quickly
because the primary file arrangement of the Intellofax card was
111`.
by subject code. Beginning September 1950a the Machine Division
started an adjunct irea File by preparing one extra card for
each main area (there was no card filed by related area), No
subject code was punched into the card. The Are File filled a
specific need at the time, when many analysts were woefully
ignorant of Korea. It continued to serve effectively in
retrieving all information on smaller areas, such as 1,he oblants
of the UBSU and the r4"vinces of China. ecause the Area Dile
grey so rapidly and was consequently useless for large areas in its
set?up without subject code punches, the Analysis Branch and
1953
the Reference Branch made an agreement, concurred in by the
M..4chihe that area cards would be punched onl: fuepf
* AnAysis BrancE7Fcfessional personnel were cal Yd by various titles:
classIfi_ors, indek..11.3, coders 1/4.-!_he most common, but the least professional),
and finally Library or document analysts.*
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040
:a ,rodtno ced
Of' decl cotion
.J j.
INTTilLOIYX ?pa ge tL
Ares Cramoved For Release 2001/08/01 : ippiglatagial300040008-9
the European satellites (except the Russan
LA CM4
Zone or Germany) -Jssa oblasts, Central (except Mexico) r>.nd Sonth
Arrierican countries, African countries, Asiatic countries.
(incl. Near, Middle a ad. Far East), Finland, Yugoslavia,,
Trieste, the Vatican, and islands (except Australia
and New Zealand).
.AnA
LI 1955 another important change was made .tet the
Area File. The classifier underlined one ,subject/area combination
considered most representative of the whole document-0 The
entire
igit c,,Irtbject code was punched into the area card,
but within a given area the cardww,-; filed only by the first
digits off: the 130.
0Ex8clud9eTfOrfattertH
GROW'
downgradino and
Approved For Release 2001/08/01 : ClMilitalreg51R000300040 ?
declaasificatian
TVIEIIGREC--page 17
Other coclo95
A proved For Release 2001/08/01 :66#1444110951R000300040008-9
3. Miscellaneous Codes
S3curity Classific,ition
With the co-apletion: although continual revision:
of the ISO 'Ind the adoption of the AMS Area Classification:
thought was also given to other necessary codes to be
punched into the IBM card for complete retrieval?
And,rmxr, issued a memorandum on 3 January 19149
,/
to "All :Elinds: Ca: establishing uniform codes to be used
on all OCD coding operatiomo The Procedure !artuals of the
lintellof= System (1910, 1954? 1959, 1960, 1967) show
the security classifications with various controls
evolved as more and more non-CIA requesters used the
system In l99 in addition to the actual security
classification codes:, then; were only two t!,..E.K-,s of
coft-%rols--US Officials Only and CIA :Tilton-al Use Only,. Over
the years: others were added to the coding pattern so that
the machines could ell2r4ate certain document; citations
11-ith controls such as Controlled Dissemination: Warning Notic
Sensitive Sources: No D iS S En Abroad: No Foreign Dissem, etc,.
b? Source Locator
In June .7198 the Library issued Library. Bulletin 'f:o.
Entitled the Locator System: if explained that the intelligence
document files in the Library had been sot up accordin to
"A" and "S" files An arbit,rar?y ivision it was de V iS d for
GROUP I
1cluded from outornotil
downgrading ond
derlasatoflon
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
L:IT:21:4.,OF--ogo 16
C woved For Release 2001/00951R000300040008-9
practical and simplified location end filing of documents.
"A" files included mainly attache' reports and State
despatches as well as CIA raw inteJligence (00-313 add SO ,
"3" files included mainly finished intelligence, intelligence
symmaries, monthly or weekly reports, end the like. The
first number in the locator was a Jvdigit code assigned to
a par bicular agency. The remaining r digits were the country and
the post for "A" type documents and branch and division oC
the
-
the agency for "S" tipeo Thus.1 05&0601 referred to ArmjleeattatW
25X6A report froift
and 0550601 referred to
Army-series
-
Armyeseries type document from the liar Last Command, PZIS.
These same designations were also used for indicating the source
of the document on the Intellofax punch card.
By 1 Jane 19li9 it was necessary to issue a second
bulletin because of numerous changes in organizational divisions
of government agencies. In the intervening year, in addition
to the "A" and "S" type categories, four more had been added:
"C":.k correspondence and Executive Registry material, "G"...?)nsic
intelligence studies, L' bibliographies and "F"..--1,ress. By
February 1950 these arbitrary tspe designations were no longer
punched in to the Intellofax cards
The igit source locators remained basically unchanged
until May 1954,When specific city or post locators for Amy,
Navy and Air attacN? reporn =e no longer considered necessary
08 i;m"m
Extludnd from cMornotie
' down 9
C10- tthiftV51R000300040 grading and
Approved For Release 2001/08/01 :
d lassifiontion
r?v_ixr--ro7n 19
C 'Ark.C1164:ii For Release 2001/08/01 MegiltiOgiOe' 0951R000300040008-9
for retrioval?? By
44A) ^
(Tr'nuiary 196 orqy th:3 4 di E it source
wa used for everythins; ecopt Ci for,3ign:9vernnnt reports and
Secret doculilnts
01- Air
02- nIA
03- Navy
04- Stato
05- Army
06- De.f.":-.,,nse in general
07-114. Other government agencies
15- Executive., Legislative and Judicial Brunches
16- T,Tonlovern..-ent
Intrnational Car ganiz onz
10,- Foreign Go-v-,-:rn:nts
? The coding schemes described in the previous pages
provided selectivity in retrieval. Requesters were always urged to
be as specific as possible on subject requests and not to ask for
too general a subject, such as Politics (the entire 100 chapter
25X6A of the ISC)-11111111 The only reason for 6 Agit ISC was
to pinpoint specific subjects, if possible. provincial breakdowns
opSSR and China helped area specialists. Requesters were also
reminded that the date df publication was punched in the IBM cardo
: Why ask for all years when only 1950 was needed? Security
classification and source specificity were part of the retrieval
picture, although not requested as often as subjects area and
date lielittions. Requesters scnetimes thought they knew the -
source of a document and.they-praved to be wrong when a rerun
was made Thrall sources, The same was often true about date.
The following is a typical request ung all the
code selectivity:
EAdz:1%1;
e f,11 j Und
,111,1;,thOn
Approved For Release 2001/08/01 : Cgt.lit#6951R000300040
'Approved For Release 2001/08/01 ? CIA-RDP84-00981R000300040008-9
4111511.1SWiliiilims
1;%13514niCt PaX,,
..pmatration of labor organizations in
during 1949-50. CIA document
Through Confidential.
(SO) only,.
card forma%
114.562
Subject 'code
(columns 1-6)
.
Country coda
(coIum.A3 7(a0)
49-50
Dato3
(colmis. 24-25)
02-0404
Bourc
(columna 13-20)
3
Security claseifieation (uolumaf,t12)
Abbreviation File
A reference tool which the classifiere found a need for as
early as January 1949 was a list of abbreviations of organizations
which appeared in intelligence docunents. A manual file of 3" x 9 cards
was established out of necessity because there was no one list of
abbreviations, particularly of a classified nature, which net the
complete need for identification. A statement of functions
of the CIA Library ih September 1950 included : "Maintain and
service a central file of abbreviation., and code names for intelli-
gence documents."
Established originally because of an indexing need, the grawirg
(49socio by 1959)
Abbreviation File was also used by the reference librarians when
published lists of abbreviations did not answer specific reference
queries. The card contained toe abbreviations the area, the title translatic
the abbreviati
the foreign title, a brief descriptive comment, and the source of
ffices throughout CIA, particularly FDD, supplied hundreds of
abbreviations and their identification to this File. A not appeared in
G1ZOUP I?
Afz7nog Wcad
OIc!er.k222i!icotion
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040
ORET
LITELLOFAX.wni
Approveestorwelease 2001/08/01 : CBLII?IlitilLiktr0951R000300040008-9
the front of the CIA Telephone Directory under services
of the CIA Library encouraging requesters to make
use of the File. In 1994 a publication was distributed
entitled "Abbreviations of U.S., and Internationa3. 25X1X7
Organizations of Intelligence Interest 9 (CD # 1).
Thereafter requests were received to publish certain segments
of area interest, such as all Russian abbreviations, but t here
had been no attempt to confirm translations or even the correct
foreign language title.
Approved For Release 2001/08/01 : C4Arkilikter951R00030004
GROUP 1
Excluded from automatic
downgrading and
ification
20
1.;LAPproved For Release 2001/08/01 : gailatiffialigiffi1R000300040008-9
Co
a
15k12A7b
Daily Rei)oe',a
"The Rr:Tercn. e CL.nter as eager to begin some kind of
indexing lisinE, machine techniques, so -while negotiations
were continuing for the development of the Inteilofax equipment
end for the construction of a unig3(eaA1as3ification scheme,
management decided to indelt the Daily Reports. This
actually became the first data base available for machin2
retrleval from the Reference Canter. Two analysts from the
Classification Unit wore as signed. this task and that manpouar
coverage continued until ind 0)ring was discontinued in
Card punching began on 19 August 1947. The index cards
contained the following information: security cl 2 ssifica1dori
pamphlet date; one edigit subject from 7jor subject
categories (Aran Tiavy, Air, Sc:ience and Geography, Domestic
Political Affairs, Foreign Political Affairs, Econonics,
Soc:7.ology. and Miscel4Wous), page
reference) intercept
headlinol transmitter) target countries. 99 countries :ith a
digit sequential code made up the' area codes. The requester
was furnished with a 1istin of the selected cards
intercept heading; page reference) uiph1nt date. a ecu r
classificatioi and areas. By the end of September 1947
14,762 cards had bean punched ad filed, by November the index
covered reports issued since
'''2gX1A7b
This methoe, cf indexing Daily Reports continued
GROUP1
Exclud,dfmmoutamatic
clowqvgdin j and
Approved For Release 2001/08/01 : 4- 0951R00030004008LV"'"' -
That:' Base
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
until January 1949, when a procedure was put, into
effect
allowed for a redesign of this particular
IBM card in order to make use of the MG subject and
area codes. Again only one subject code-vas permitted. The
name of a prominent individual was entered in the first
15 spaces of the which was restricted to.110 spaces?
A ;(,rear later and the CIA Library jointly
prepared for inclusion in the front cover of all restricted
issues a short announcement advertising the available
indexing facilities? Based on recurring requirements from
the Library sent out
certain offices, particularly ORE/0E1 and OSI?Atyped
lists of pertinent titles to these offices every. week.
early January 1952. the Library a dvised, the dis-
continuance of the indexing of the Daily Reports for 4tag66P-62---
reasons: Requests averaged only 10 a month plus four -
recurring requests; and with indexing restricted to one
subject per article because of workload in the Analysis Branch
and in the achine Division, adequate cross :,-..eferencs to
cover all subjects within an article could not be Inadee
The Chiefs Analysis Branch in a memo to the CIA Librarian
stated that the two analysts thus released would be able .to
devote full time to help reduce the Intellofax backlog of
several thousand documents
GROUP 1
Exctuded -horn automatic
downgrading and
d,dass;F;cation
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
1J .1 'LUZ x.--page 27
Dat-1 Bos:;:a r 1
Approved For Release 2001/08/01 : Cl - - 1R000300040008-9
Dr. Andraos concurred bccarse of severe 1552 mnpodor
outs and because the OCD Rczisters picked up the
found
personaliLy and ildqstrial plant infomaion\in The Daily
eports. on 6 F ebruzrY the Library disconinued thP cioding
of all radio broadcast information. Although the IBM
cards were retired to Record5Center
the Library retained
a master printed form of all the coding effort.
The issue of tbhiaiiiability of mwestablishing
a machine incle to the Markt Reports was raised
periodically. (See Library Consultants Report of l957
and Berner Contract of 195W
20 Zarlir Intelloiax Covera-yo
With the publication of the ISO in March 191t3
it was possible to start inderirm n earnest. The first
efforts were confined to oqp3 reports. One Transmittal Sheet
was prepared for each docunent: jt contained a bibliographic
statement: an abstract of the contents and pertinent
codes Until the Central Index had typing_.pc-:_esonnel and
reproducing equipment to type and reproduce abstracts on the
tabulating cards: only the punched data appeared on The ?
IBM card' the Tra it a1 Sheets were filed in t ha Library.
Plans called for the receipt of ;000e. ocuments
a day. EYperience already showed that a classifier could
abstract onIr 33 documents a days Becker noted in dune 1948-
18gLiiititifu?
1 downgrodinq une1
Exchocled fmm outomH
GROLIP1
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
rt:TELLNAK--page 23
ilasApproved For Release 2001/08/01 : 3iiiikill064141951R000300040008-9
25X1A9a
that, a T/0 of 20 professionals in the. Analysis Section
would not provide adequate manpower to abstract every document,
In 7?-1.(nber the current intake was between 400 and 500
:7.12,'em,s a day. The 1948 backlog of a ?proximately 12,000
SO (predecessor of CS documents from the Clandost:In
Services) and 3000 other CIA reports was decreasing
37-/
150 items par day* Of the backlog of non-CIA
reports it was osti'llated tbathe 154,000 itelas would
nob warrant indexing because of conton'o. The unclassified
and restricted doaunants for 1948 were indexed by Special
frojects # 1 ("the pool"), 1Jocusrnts issued in 1946 and
19)17 TaPc proceqqed but only thode of priority areas of
interest* Becker stllted t"-iat it Tauld eppearossible that
"we can get a 1 January 1949 target for providing 6aily
ta:b-fax srvice." And this did occur,
in report to the Assistant Librarian Oil 9
March 3949, gave the follovinE status report
of af
In;-,ollofax coverage'
.-
a KL1 'A" type reports were currently indexed since
Selitr-16)er 3.9118
b. 115" type docuirents were selectively indexed, such
as ail state cira roc o.ets, and Top Se,-..'ret reports,.
0 P11 C Cfr're SpOr .?;r1C.0"C -{T.Zda. Reg:1st:7 nurriber.
do All bib1-1.og-rai)1.1.1.E-,s on file in the
e P'11 Joan ib ors
GROUP 1
Waded from autarnatic
dawagradino and
L _ declassification
Approved For Release 2001/08/01 : CR-1140P84-"-00951R000300040
ifiTPULLOFAX?page
Data BasaApproved For Release 2001/08/01 : citrfletel0951R000300040008-9
Heavy-backlogs frequently required stringent measures
that affected coverage/ For 4 Months in 1949 unclassified
State Department despatohes were not indexed. Ilde Was 25X1A6a
briefly expanded to include any document from or
African and Latin American posts. No effort was rade later to
fill this void.
Document coverage rose from 46,681 documents in 1948
to 227,106 in 1950 or a total of 41-11,0329 documents indexed
22/
into the Intellofax Systen the first 3 years There are no
comparable figures available on the number of Intellofaz
requests received in this sithe fcetiod. From 1951 through 13
aqf
requests from CIA and non,TCIA usrs averaged Al monthly.
FOROUP 1
Excluded f,-um automatic]
_d
Approved For Release 2001/08/01 :1111t111,141iilia51R000300040
INTELLOFAX--T,age 25
- Node>:
r N
Approved For Release 2001/08/01 : C - - 0951R000300040008-9
NODEX
Early in the Inc ering processiztg it became apparent
that certain documents dealt with information 1.?7-.??t was or!
little or no intelligence value for retrieval purposes.
Th., term "NODE" was coined to repreuent those occu.creInts
would not be inde:-:ed into the Intellofax 3ysterric,
Originally these were documents of a purely administrative
nr,ture. AS the System grew, however, more subjective
judgment as to the value of certain information for I.Itoli.o.fax
retrieval purposes was exercised and the list Ci NODEX items
grew and changed. In some cases, It was subject information
was rejecteth in other cases, it as an entire series.
There?ww3--no ent the rt,t,eipt
4he ency?Trdg'rr'6---
want tt--oe---thegl.? The whole question of what should be nodexed
was stetar much debated throughout the entire Ini,ellofox history.
No two researchers agreed/and reach criticism was levied on the
System because of certain NODEX decisions.
The selection out criteria irr-uhe t-ar1-7--dayu---of__';Uad
entirely
fellAupon the class if ier,v7.11.1c---uou3.-A----so- ynnk
a 40,citriient enveropE--.
(11'he Incoming and Dispatch Unit of t he Library ;--3,00n- recotnized
certainseries:, such as Array I'lhois Who eporbs. These were
btchod separately and Cid not even come to the attention
Ery I 4 Ili
GROUP 1
?
Sxcluded from automatic
downgrading and
dIdossIfication
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
uITTELLC;FAXage 26
cxf.'ex Approved For Release 2001/08/01 :41,1ppliR1Wifi951R000300040008-9
- of the classifier?
would It lrce Files for every
docume
an ab
nura itle a.
v.-41...5_:rniar_e_ci so that thgsre
ver, coat 'Lid only
arce2cc nunb
Etti on, T1
,--
iated
CO
tion
created p- in search ng thro
the c., ,unent
The early 2.95Q .NODEX Standards included such
topics or series as:
a. Purely administrative maje-,ers
be Consular or commercial furictions (replies to
complaints of Americans about lack of sex-vice)
ce Notification of change in sect,-.'ity classilication
de Agendas of various inLernational committees
ea Order of battle (considered a military -f-eq-ponsib:lity
re Transmittals of enclosures not; attached and
not Cesc ribed adequately enough for indexing
ge 1-m_iustrial Car. ?. File (TM) reports giving
primarily plant data
he Who Is Who reports
ie Joint Weekas (considered cables)
Out of 17,367 documen. .rirj.cessed in January 19512
ow tt?A9(.04
1125 or/et f t he total 1,
GROU' 1
Excluded from automatic
- downgrading and
declasscation
Approved For Release 2001/08/01 : C - - 951R000300040 -
_PIPELLOFAX- cage 27
Nodex
Approved For Release 2001/08/01 TIMIWOM1R000300040008-9
A printed list entitled "N62-6X Standards from Start of
Intellofax System to July 1966
ePar6sntlie-s#* is indicative of the colorful history of the
NO= progri (see attachvent) Throughout the years, classifiers
and reference librarians were allws reminded that thl. list
was
-
Was only a guide and that a docugrnt should not exc.:laded fr ra
Intellofax subject control because some items within the
St.,ndards appeared in it.
(For microfilming of NODEXES see page 43)
GROUP I
Excluded horn automatic
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
oioruim
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
Coblns
OCD mam-zement clr7y in the Intellofax history
deci.led t'wt cables end telegrams would no: be considered
a l'ort of the central reference system for subject/ 1.ea
retrieval There was no question as to their current
1nt,q2iz,ence value, but 00 did not consider tem
vital to retrospective searching and therefore felt It
was ureconamical to index the enormous cable flowo
This philosophy carried .throughout the years of the
IntLllof,tx Sysem, although there was a brief flurry
of L cable induxine, experiment in l963/164 (see )
OECRET
dodop
EX*1401,03 r qviomotk
4,elgroai'njonti
de0A51.ificqtion
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
INIL,A,+.A.)rit&.- -page ey
sAIL110114444Womms
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
INtellofax Procedures
10 Ta Btch Sjotem
August 19/6 COD management, decided that a reviewn -
of the existing and contomplabed document hLmdling pz-o-
cedures should be made pith a view to determining if
such procedures could be modifio -0 e-,Teditu the distribution,
classification end indexing of docl-menbs pending the fu1fill-
25X1A9a
meat of T/0 reluirements* of Administration
and Management worked closely in setting 25X1A2a
up a detailed procedure for the use of multilith mats
in controlling and indexing intelligence documents, This
vas the beginning of the so-called "batch" systom,which was
modified mew -thins during the next 19 years, A batch
usually contained in one envelope 15-20 documents of like
source (This was similar to the organization of the
The classifiers were
dissemination function in the Liaison Divisioa)nat arganisadd by
source breakdown until Apri3, 1952.)
Put into apration In ecember 1948, the system
included the complete gocessing cycle of dissemination,
distribution, indexing, key punching and final filing of the
documents in the Library document files, On one multilith
matuith a preprinted distribution la&l.er for dissomimtion
points within the Agency wastyped bibliographic information
(sourfe, ,late? title, security class'ficatior? etc);
this mat was the bafi-a, for the preparation of contr(f -1!ps
,.? ? / ?
( a cut ci carld.;) to be attacd to each Cistributed copy of a
One control slip also became a source card (see Page 45)
L;co a'Aachnient for 1.;,,,,prrocedu-ccs
of 13 Ide-2,,rbor 14Y106ikmilills
GROUP 1
Excludgd (corn automatic
downgrading and
declassification
Approved 00
ForRelease1 1/08/01 : CIA-RDP84-00951R000300040008-9
2'
J. J:
Bvtch 6ystem IM,4**CiI1
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
The classifier used one of these control slips -or
crds-on ,Thich to iaritc the necessary codes for
ultimate key punchingc The typist prepared anothc,r
nuitilith mat to '..)e airied vith the punched. lBti carat.
Tho resulting Intel:Wax card contained fielr's on thc
loft for the cod...-s; on the right end of the card was
the printed bibliob-raphic information vhich could
be easily read'otT the nal:cc' e7e. Thiq duplicate
was
preparalo'l or naltilith glats continued 1.,htil
i96 hen a revised batch system eliminated the preJaration
of control cards for distribution pa/703es0
In SPaptEbcr 1919, eJ.ch r!sr sos ci.sq.ned
an Intellofn% stamp bearing his individual nuglx7,r, Used
instead of the classlficr's initials; the stamp :fiL,e(1
in th-,-e places: (I) on the face of the docwirn'. to
iudicat'7 tht inding had been completed; (2) on the control
card for the coes; so :t1-t key punchers could gaestion a
cl2ssIfjPY necessary; end (3) on the -Ich Control Sheet
141Ach stayed ulth the group of documents tnrogs the various
OC CS S iLIZ steps 0
A revie-4 of the first Procedure 7!_tni.?0.7.1s written
in 1248nnd 1949 for the machine operaLions and for
the indexing of rocurrionts shows the intyic7cies of the
IntelLofax Sjsteil as ib developed. A procedure ha'1.
*Se sample con:-,rcl slip
*
See smple card
GROUP 1
Excluded from automatic
downgrading and
dnclos,ificution
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
?EGRET
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
to be vo.-:.tten for every e:certion,
For example", ,cr--17srrn.c-7-.E.L.311 hxtra IBA cards
were printed for a nuiaber of offices- Top Secret Control
in order to set up its cun source card file; Contact Office
the Ini-,ellofax card
for eve7,-y CO-3 document coded so that 4,4s-could be retched
with OOs CYCTil con'czct card (and this procedure continued
until 1967); Reference Branch of the Libr.y for every
Finished Intelligence and. Basic Intelligence doothheni-,
for setting up a cumulative index subject, area
and title (this stopped in 1953); and CRP; and CSI offices
were engaged in the abstract program (see pa,ge3
A special procedure was written iCT loan, documents
had to be microfilmed, if more than 14 subject codes
appeared on the controlii slip, the classifier 1-)rote n:JAT3IP on
tht. Batch Control Sheet opposite the CIA control nurioer in
the "coded" column to indicate that dditionol Intollofax C,qr6.3
were needed.
As the years progressed, the system became more involved
and prourdltres were constantly revised and hopefully improved?
The Machine Division and the library personnel worked
hand-in-hand In developing 1),.:?tter and s-4?er m-3t1,odL? of
rrocessing the document I:1o,, taw and in tang caro of users needs;
* See sample Batch Control Sheet
OI OftliL
GROUP 1
Exclueed from automatic
downgrading and
cloclossifi cat on
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
Y=LTOFIX..--page 32
Flash
Approved For Release 2001/08/01 : CIMIK1~815951R000300040008-9
2.. "Flash"
indoxilt economy .4.4.1-ttr developed !,:raste so-
called "Flash" proceCiure, By January 1942 classifiers
recognized that there was a iz/able segment of documents
covered the same subject matter each time and -
were issued periodically? "Quarterly Mily_tary
"Weekly List of Infectious Diseasee", "Semi-Annual
Railroad Statistics" were a few examples. It was a
waste of indexing and key punching effort to coda_
these documents every time they appeared? The
I,
such a documbnt was encounterodl,dt was coded and
abstracted in a table of contents manner; the intollofax
card bore the usual bibliographic statement but without
report numL-)r or date? The word "Flash" eeered on
the IBM card. A master Transmittal Sheet on which the
abstract was prepared was filed in the "Glash" book
in the Analysis Bramh? 'Ivery time a similar documeat
was received, its report number and date were entered
into the FFlash" book on the Tramsmittal 3hect0 When
subject retrieval request turned up the original "lash"
card, the librarian or classifier knew that he must
refer to the "Flash" book to find all the documents which had
been publised and rece3ved0 This "Flash" record was the
only means of determining documeht numberJ in order to
retrieve th2 rnteral the document file
GROUP7
Excludedfrmcwomatic
dcowlemcNngand
Approved For Release 2001/08/01: CIA-Kuv84-00951R0003000
SE***611
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
Although lists of "Flashedrt reports were
periodically distributed: to the classifiers-4 and
the lists grew-4 the classifiers were alwa7,rs
reminded to glance di, the current documents to guarantee
they were sufficiently covered by the codes orig-inally
established. Changs in the ISO as well as wider subje.c.%
coverage in the s cries did. necessitate added ceded.
The "Flash" system finally outt-,lived its usefulness
Irregularity of Tssuance of certain series, the manual
labor involved in maintainingthe ttFlashrt book
(for a period one cow was maintained in the Reference Branch
also) 1 the nuisance of not having a source card for every'
document, and the advent of the revised Intellofe:Ic
irt,k;1661.4
system?all combined to its demise in 1960?
dUCIPAC?111"
GROUP 1
Excluded from automatic
downgrading and
declassification
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
,
pp,
Alljj riCt rogram .71TTIET
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
3. The Aboract:i_nz Pro,cam
Dr,, Andre-Jo as v.ry proud of his central reference
4
facilities and by late 1939 he fel:, that CCD was on
the track of a "fal- more e5?fective solution to
providing analysts with quick retrieval Of -.:ntelligLnce
information than has ever, anytihereT, been a chieved
before ." He devoted 32 pages to a dafiniti-ve uesc11,7 tion
of th3 Inellofax System, entitling his monograph
"Central leference Facilities. Status (1 liovonibe.: 1949)
and Object:Ives." (He harl prepared this paper at the
request of Chief,CJAPS for the :1 ni_ormt ion, of all of .ac')f-:, )-22/
The Table of Contents is indicate of the complete
coverase of his subject:
Summary
Glossary (of Intellofax terms)
Magnitude of the Problem
Classification Scheme
Intellofax 4istem
Index Files
Abstract Files
Highly Bps cialized Reference Problems
In his usual dynamic approach, Andrews had
briefed specialists in ORE on I-A-lc, potentialities of the
Inteflofaw. System and how analyst rerticipation in the growth
of the file would benefit the system and therefore 412
agency. In other words, he 1, as asking analysts
if they would lilse to contribute coded abstracts to the
f 13 e He stated that or'ly the specialist could decide
GROIN 1
Extluded from aotomotie
downyrud7n5 and
declossincolion
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
st,roo bing
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
25X6A
whiohittetthe important doe rents bearing upon
his field of specialty End on. he specialist, coviLa
vreite competent abstracts of such d ocummtst His
fit approach was to the MA Division of ORE
because he knew the Division Chief well, IN January
25X6A
of 1949 the began 3.ecting and. abstracting
25X6A
the most Laportant documenbs on The assumption
ME
was that if t he sL?rsbem could be made to work satis-
factorily for one area,j then it could possibly be
extended to specialists on other areas, 051 also
coTrraenced operations on a trial basis in Februarys
and the Gre
desk of ORE in March. (History repeats
25X6A
itseLU During 1972/73 the GTI area
and
of CCI is providing input to CRSts AEC/1-3
compuber system on an Er,cperimental basis.) OCD Ineanhile
continued to write abstracts of a table of contents type
for publications covering a wide variety of subj'3)cts
and areas? On pages 23 and 25 of the Intellofax
Dro Andrews provided samples of CCR J OP:o and 031
abstract cardso so-called contributor cods was punched into
column 21 of the III card so that, if a specialist ever
wanted to retrieve only his or abstracted mat-Tian he
could do ao,
The two desks in ORS providing thcne abstracts
OD:MET
rGROUP I
Excluded from automatic
downaroding and
declossification
,
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
11...411; C T OFAX- ge 36
Abstrcoting
Approved For Release 2001/08/01 : M-IITIDP84-00951R000300040008-9
told 1J2v Andrews that the abstract files were proving
exceedingly useful and that the process of writing
abstracts had in itself yielded a nyeler of unforeseen
W%-?4444.A-0
but highly valuable by-productsb 54a0,4-a-s4 time
saved in pradoc:ing weekly and mcnthJr summaries0
carbon copy of the aostract or Transmittal Sheet
was flied in folders in a strictly chronological order
t4
and provided the desk chief a fully documented hiqtory
of day-to-day events in the two couni,ries, raining of
junior professionals was improved) speeded up and
thoroughly locked into the production system by the assignment
of writing abstracts
echnically difficult reports
no longer went into the Fhold" basketbutwere analyzed,
researched and abstracted, ,ile space was sav-d because
the abstract could replace the original report.
Much as he had fostered and approved of the abstracting
program, Dr. Andrews became concerned about the rapid growth
of the abstract files in Intellofax. By the End of the year
there were 16,047 abstracts (002-28475 OaE;6245;0SI*,6955)
and they growing at El staggering rate, He warnedthat
each ettension of the system to a new group of speciali-ts
involved difficult adjustments, revisions and expansions in the
classirication scl,ero, required more typists miu could be
accompli9hed only by da3r-t,2-day hare, .Torh, How trueI
Sarni:W*1mm
GROUP I
Excluded frorn autornotk
downgrading and
dectotailication
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
.P:b3tracting, Proo'ara
? Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
25X1A9a
spent a large Tiro-portion of her time in
1949 end T-Torking on changes and expansions of the ISO
to satisfy the needs of these new contributors Area codes for
three of the .,1-ecir Eastern countries were expanded to take
care of provincial divisions? This later created some problems
because they were never used for retrieval in the Intellofa:67
SyStelac
Dr. Androus conolueed the iguana/5r to his study wih
these pertinent words:
Because of the 7.clectivity being exercised in
building up the abstract files, we are forging a
tool which in years ahead will enable us to drain
off from the Library thode documente which are
of seaut importance, thus. making room for
the current inflow and ensuring that reports of
real impertance are held available? It is quite
possible that the central. reference system being built
by CR 1-7111. ultimate Z prove the most important
central intelligence service which the Agency provides,
The abstrcting program mushroomed f rem its beginnings
in 1949 reaching
in the eavly 19501s0 The
25X6A of MA joined the program in l950as
did Lgricultureia Division, I1-950---ORIt-contribute.d
.21,d
_16,55e?ab-strdots;-.0520;la6.4 some of these desks even .ob up
their filin, fl-Ftells
o the TSC subject breal-f,o-;:-Is4
onpup
Lecluded from autom,tic
downgrading and
declassification
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
25X6A
Abstraot-in:; Program
- Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
; ?
4
, ?Tt .^4
VVilljtC :.\ '
," r) )
. ( "'
? e ? '
It' di IU
tb4
r 1A-
(A/
Other divisions or branches. began to contribute in 1951
on the encourageent of the Assistant Tdrecter,v.
whew as interested in space sav Ingo ORR colitribl.2.ted
16558 abstracts in 1950 and 10$695 in 195:4
25X1 A9a
OSI coutrThutions
25i/
amounted. to 20186 in 1950 and 20715 in 19510?
Hvi:ove.0.1 tite- doT,In,:ard .tx7,3 began in 1952/53 as
specialists in ORR and OSI were interested in spend:InE; their
P
time learnin_ ho u to use the IC correctly for in-depth
relviv-ttA
indexing ;-L3I/Physio s and 0S/Ci eine rwe heavy
41..0k
eonLribntors,*: but finally 0,03,706. in 795h and 2955 :'es.pectiT:ly
ORR/Shipbuil TLn
arl/
zzcat5E-
el, last oRn conponent to et (-.r) innut
OCD management had altars hoped that these
contributions
would preclude the indexing of the. same docu.ents by u.o.e
OUP. classifiers,,hi 1::1s never possible, for the specialists
fregnently dxtracted and coded__ only that part of the
document that interested theme There was never any guarantee
that the entire document was covered So- duplication continued;
both
IBM cards would turn up on a machine
run if the sal:ie codes
had leen used (and there was no guarantee of this., eit1ier.)0
7.:b2arian in
In. surfo. a CP., the oaD classifier or t bc
screening thfl cards before an Intellofax tape was made would
pull put the nonbstracted, card as beLlw less mealdnEfulo
STPA9WilaTi.
GROUP 1
EL:cludod From aulonlaticl
downgdin
rag n nd
declassification
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
T=T.T.1,2JKpage 39
410(011,11111"a
Af3-11rthierd.;Fkir(Ralease 2001/08/01 : CIA-RDP84-00951R000300040008-9
-ifu).Puv
A groat danger to the Id:,e:.1o."aY Liysi,c,m later
0, ourfaced th, re3enrch caiceJ had E?to,-Ted the
P1)0Llactinz, r1oLr,-4 Referenc,s turned up could
not, bl:t retrieved ,arly=kmmgmr, sneh as articles in Russia
scientirie jeunrals indexed by 031 and lAa- tbroon cr,Ty.
7o limit as to F;oyrce notarial had been p1u,-0 on the
Jpoc4...listo rci.,Mt even wtto extract or pre:prxe an
ab9trLet fro; q the Ive-r- York Tirer, In the ni'3-1960'6
(amet dai,,-; not conf:rmed) all rdi*1 cards with the contributor
code wer pulled and destroyed for they added nothing to thc.,
AINI/
Intellofal-t: file, but?\.-ithr created rPrie--ni difficals.,
OCCitur
OnOUP I
Excluded from automatic
dcv.ngrading and
thclossiticat!on
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
INTELLOM--page
Approved For Release 2001/08/01 : - -00951R000300040008-9
Microfilming
E. The Microfilm Program
One af OCDts problems was that of keeping the Library
operating at a maximum peak of efficiency. The Intellofez
System had made available a greater volume of library document
references to a larger number of requesters than had ever
before been possible. The increased output of the System
had resulted in a corresponding increase in requests for the
documentary materials referenced. In wishing to offer maxim=
service to all offices, the Library was faced with the &Mama
of coalescing two variations in point of view as to these
services. On the one hand some of the Library users insisted
on an in7io1ate set of d o cuments in the Library at all times.
On the other hand, some insisted on the availability of
library materials to their respective offices On demand.
In answering this criticism by an ORR analystibecker wrote:
We clearly recognize the need for ensuring the
availability of a master set of documents;
bowever, keeping an orie.nal document collection
poses problems of filing, space, circulation
and reference which are almost vverwhelming?
With 617,562 intelligence documents on file (but not
all indexed) in the Lterary by September 1950, space had also
become a serious problem. There was no doubt in anyone's
mind that microfilming was urgently needed.
Approved For Release 2001/08/01 :ffilifigillilipili00951R0003000
GROUP 1
Excluded from automatic
ociraccad
(f Approved For Release 2 01/08/ sI : CIA-RDP84-00951R000300040008-9
Geht/X0 )1'
f(,>415/)
3
a/A-')Iit;i7 S;)I
A-P4 tf"-
(446t4-1'
IApproved For Release 2 01/08/
I: CIA-RDP84-00951R000300040008-9
25X1A9
?mrrenau--page hi
Approved For Release 2001/08/01 : CIA-RDP84-1=4X040008-9
lacrotilming
In March. 1950 the Library began experimenting with a
microfilm and print procedure and by mid-1953. it beg*
to microfilm all single-copy material on 3ara, reel film.
The Machine Division and the Library worked closely
together to develop the best sort of system to solve the 5Library
2 1 9a
document storage and retrieval problem. In January 19
25X1A9a
his deputy, examined equipment at
25X1A5a1
wherein microfilm was
mounted into an IBM aperture or window card. This system
allowed each document that was microfilmed to become a separate
entity in itself aad not just part of a reel/ as was generally the
case in most microfilm applications up to that time.
In October 1951)Secker told Dr. Andrews that the
problem was urgent and ite- proposed that the Library microfilm
all incoming documents, keeping a copy of the document as well as
the microfilm. The latter Would be available at all times both
for viewing purposed and for reproduction in cases where the
requester wished to retain a core on 19 December 1951 'tee Proijeet
keview Committee rathorized-nmierofi3mdne-ef-allAgnifieant incoming
25X1A1a intelligence docvnent&1 d approved funds An the amount of
25X1A 1 a
for the iial purchase of equipment a nd in the amatmt of 1M for
the annual costs of personal aervioes and' 'suppliiii.?912-4
and
25X1A5a1
Approved For Release 2001/08/01 : CIA-RDP84-1661,1119W0040008-9
GROUP 1
lEneudnd horn automcdial
trownstading and
deda.ssiRcation
4.1a
l':iler?2'41Med For Release 2001/08/01 : Sli141100613151R000300040008-9
25X1A9a
25X1A9a
e
IMMIPMME)ecartm juASgtaNfircANtatilitiiiiiileilatilin&fioe ef
almAral Serices.__
Organizatiom and Methods Serr sarveyed-06 prop osed OCD 25X1A9a
Nicrofilm ProjeoYand-submitted a-final report on 15 October 19537 1111111111111
bc.tiA ,\
:She-recommended that the personnel required for operation of the system
bp' On the OCD T/0 and under OGD supervisio
of-General-Sant-efts? Aftortling-to
,mould achieve by 1956
r-thext-waderlthe-Offiee
25X1A9a
the proposed system
$: immediate -film inspection,-
immediate customer service on the entire colleetions.eliminate tracing of
documents and repeat borrowings from other agencies, about 90% econmqVin
space And filing equipment; better utilization and conservation of personnel,
4Alat'pernit reduction in size of researchers' filee and mail-expedite the
chain routing of documents 12
11t40#14i.T.
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300
GROUP I
Excluded from automatic
down9rading and
declassific flan
IX -pg 42
r 1
mierazaTOroVed For Release 2001/08/01 : - -00951R000300040008-9
Research and testing_ continued throughout 19513 1952
27/
and 19530 At a CIA Bidet hearing in 1953 Dr. Andrews stated that
the Machine Division's wine job wne to keep pace with new
dervalopmento all over the country but that the most important
research then at hand was to develop a microfilm processor theo,
in conjunction with the lectellofax equipment, would ensure that the
Library T7ould give to the analyst the rocuments which he asked for.
iicroI.iming of incoming documents began officially
in March l9' %e following equipment had-I:leen-ordered and-was
on_lantle-two microfilm cameras for the purpose of making
initial microfilm reels, three
one Nicrotonics Elba Printer for coring the original reels,
(one copy of the reel was placed In the Vital Docurents-Refoilitory),
three
"mounters' to cut the microfilm reels and install the fra.mes in
the aperture cards, rine Photostat Printer-Processor to make positive
prints from the aperture cards,
The aperture card As a punch card which had information identifying
25X1A5a1
25X1A5a1
-tbe microfilmed document printed across the tcp and 16mm. film images .
of an intellof axed dnument mounted in apertures (openinga4 on the right-
7)
hand side, , Aperture cards varied in that a single aperture might
centain one, two, three or four apertures Each aperture contained
a inOcizara of too film images, each lage being one page of a clocemento
The basic procedures of microfilming documents, did not ethange materially
until 1966 when 35en. film was used for the preparation of aperture Care3
i'nstead of 16m. filti, Detailed procedures were outlined in
. Appralliiiert6PFCAea44 iOVMA-It' C&W13n4lcdargfgagOlRa(80009-3) ' dated Dzcerebar 19590
TAT7,110FAX..-pao,
Miorofil.ming
A screening committee .of librarians and machine people
Stfi2*(1992...p1/443[IFYib493119514
0030004904UTOcuments at
-
the intake point and to work up proper procedures* The first
documents idcrofilmedwere.State Department. despatches These were -
followed?by Air Force reports (from an original ozalid copy, which
had to be returned to the Air Force within 48 hours) in April; by
Army and. Hamr reports in July; and by the last segment?CIA-reports.
in September?
The aperture cards were filed in the Circulation
Dx.anch of the Library by control number assigned to the
document* 35mn? reel filmwas used for documents over 50 pages
in length? Bulky and oversized documents were not microfilmed*
The remainder of the documents were photographed on 16mme filiu -
(Else page 46 for microfilm designators of control numbers on
sovrce cards0)
- With full-scale microfilming in effect, the
Library and the Machine Division soon decided latein 1954 not t*
microfilm NOD EX documents because their contents did not meet
indexing standaedse In April 1955 this decision was amended so.
the microfilming would occur only for thole MUM that were
single cows required -further routings contained enclosures or
were of CIA origin, thus ensuring an inviokate oopy in the Libraryo
In the step-by-step proces ding or batch procedures
est4blished for the flow of most documents, microfilming occurred
aCtor indexing, so that NODEX determination could be Made first?
This had one big disadvantage in that the microfilm of the document
was not on file until after all- other provessing had been completed*
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
INTELLOFAX.-page 44
. Approved For Release 2001/08/01 : CIA**14144951R000300040008-9
Microf
Print service from, the aperture film becaxae the responsibility
of the. Circulation Branch. Any equipment developments or problems
were the responsibility of the Machine Division, such as experimenting
with improred aperture card positioners for FiImsort viewers and
with methods of printing documents from t he microfilm viewers.
OROLT 1--
Ewbckd from ourdmotic
Approved For Release 2001/08/017IMMW41210951R0003000 oingtrazd
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
Figure 13
Aperture Card From
the Aperture Card File
0
4.4
0
(S r.-1 c?-? IVA ,,, .^ 1- in^ ^ CO
0 r.Z. K 01 0 41. LC, CO
0- ? 0
la, CO
CP
0
0 it K ay 0, t.t.A sa- En co
? ca re, O. In ea
.41 co
te,
.0 In
Z K Ka 0 .0 41,171.
re,V-? .02.
_ -
t..'r CY 0 rciA 0 CO
- CO
CO
?.
CO
1,1
a
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040008-9
- 71 - C-O-N-F-I-D-E-H-T- T- r
Approved For Release 2001/08/ Il : CIA-RDP84-00951R000300040008-9
?TNT
crfI 7/itt j/u /4,6rV
I -24 6L'Ir
)1 M(4:A 14611 6Y7A/
Approved For Release 201/08/
1: CIA-RDP84-00951R000300040008-9
INTETMFAX--pags 45
Approved For Release 2001/08/01 :400iRtipt00951R000300040008-9
, Cord
9012100 cards
F. T1e Source Card File
An important by-eproduoi; of the Intellofax
System was the eplabliahment of a source card file* The
Library early recognized the need for a card catalog
of document sources, similar to the author file
in a book catalog* In the first coding efforts of
1948 the classifier wrote t he codes on a 3 x 5
form on which the typists had typed the bibliographic
data* This. so-called 35-2 (044-GaLled from the form
number) was filed in the Library by source after the
completion of key pundhing0 With the Anauguration
of the Batch System and the use of multilith mats
for the preparation of the bibliographic data onto
the IBM card, the Machine Division provided the library
with 3 x 5 *wet,
. nese "cut" Intellofax cards were used as
:* *
A
until 196riehen the Li:beery 4reed to accept from the
28
Machine Division a punched Intallofax cardrthe
pdeantage presumably being that these source cards,
could be sorted daily by machines.
The source card file served several purposes:
(1) inventory of document holdinge) (2) identification ei
a document/and (3) location of a document* Requester*
looking or a specific document often did not have the
32/
* See page 29
** Discussion had bens' Mirvinitigialtiet
Approved For Remase zOcuus Ou931 R00030004
GROU9 I
Excluded from automatic
downgrading and
itication
1-arnAP035.veslarger lease 2001/08/01 : CIA-RDP84-00951R000300040008-9
solisrufm-
card
&,cument lumiberbut might know the source and/or year.
The source cards uere filed by source, year of
specific post or agency breakdown and document number.
Veterakm
A brief. tttle_description of the enclosure Aso whether
was received or not received, microfilmed or not
microfilmed, appeared on the card. After the librarian
in the Circulation Brandh had identified the document,
the
she could then find it in the files- either in,hard copy
files maintained in the samesequenee as the source card
file or after 1954 on microfilm. The approach to the
miorofilra was only through the document control number
44a-
which appeared on the source card:
"D" control number--- an 16 rm. aperture card
"C" control number?. on 35 mm. reel
RV" control number-- not microfilmed and in hard copy
-{ke"
The source cardifidarwas Prepared for NODS
contained only an abbreviated bibliographic entrys 1.e.,
sources document numbers date and security classification.
The title and country were not entered. This abbreviated
ndation saved typing timeebut created prOblems for the
Circulation Branch librarians who searched the Source Card
File for document identification.
Approved For Release 2001/08/01 : CIA-RDP84-00951R000300040
A
_gRotn, 1
xc ?dad from automatic
downgrading and
declassification
Approved For Release 2001/08/01 : ClAilagaN1R000300040008-9
Library
Maintain and service a central filo of abbreviations and code names of
intelligence documents
( Statement of Functions 20 Sept 50
from Ex AD/CD to Management Officer)
In Felder #00Agii/X1X1/0 Box 59-875/1
OCD History 1947--
GROUP 1
Excluded from automatic
downgrading and
declassification
--Api3rovcd lor-Release-2004/0 . IA-R131;t64-8095-1Re00880840008--9---