PROGRAM FOR LIBRARY CONSULTANTS MAY 12-14, 1958

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP82T00271R000100120004-6
• Release Decision: RIPPUB
• Original Classification: U
• Document Page Count: 260
• Document Creation Date: December 22, 2016
• Document Release Date: January 4, 2011
• Sequence Number: 4
• Publication Date: May 8, 1958
• Content Type: OUTLINE

Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
(_... '' 6 ",,) .
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-
r
8 Play 1958
C 1 1 N T R A L I N T ELT, IGEI4C E AGEI C Y
OFFICE OF CEPNTRAI, REFERENCE
PIZOGRAt-i FOR LIBRARY CONSULTANTS
MAY 12-142 1958
FOil OFFICIAL USE ONLY
I. LOCATION
Office of the SA/AD/CR - Room 1053 M Building
II. MATERIAI,
Task Team Reports
Consultants' Survey Report
Consultants: 'orking Files
AHIP Documents
CRAG Documents
Status Report on OCR
Task Team Roster
-~ CCR Monthly Reports
Report
III" 1ST0 SESSION MAY 12
a. AD ; s 1 elcome
b,, Outline of Past Year?s Developments:
10
2.
3?
It.
Task Team Organization and Objectives
Establishment of CRAG
CO]2IAC Aims
Training Library Transfer
Outline of Consultants, Immediate Task
1~
2.
3',
It.
Review of Task 'T'eam Reports
I:i.scussion with AI?/CI% and selected Task Team Chairmen
Agreement on procedure
i?lode of reporting reactions to OCR action taken to date
d? Heading of. Trask Team Reports
FOR 01-FICIAL USE ONLY
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
Next 1 Page(s) In Document Denied
Iq
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82TOO271 ROOO1OO12OOO4-6
CrLAG :12.58
10 I'/jay
C ENT I AI, NTI L7, IOIi NC E A G E NC
CENT-W, I II'I?RE CE AJUVISCTI GROUP
Sum ofZ ask Team Eva:l.uatio;is of the Findings
racy ~%onsa s
1. loath your cooperation,, during the last few months., -task teems
made up of personnel from your offices and mine have been engaged in a
review of the operations of the CIA Library and related activities.
These team efforts have been focused on the recommendations ramie by the
OCR Library Consultants a copy of whose report I sent you some time ago.
As you Imo: , I was anxious to proceed in this manner because certain of
the proposals made were so basic that their adoption would have reversed
* philosophy and approach to intelligence documentation built itp over
a period of eleven years I did not feel that the amount of izivestiga,;ion
permitted by the relatively limited amount of time available to the Cori-
sulta_nts was sufficient to act on their findings alone.
2. The sixteen task teams have noc, co.r_pleted their work and have
sent me their reports. These are of course the findings of the individual
teams and, despite some coordination bets eea groups they contain ~ .Y .n some
FurtaleP
3 Attache(3 for your information and., in some cases. future C TAG
action, is a recapitulation of the problems posed and the summ ry fi 1C?__n`s
of the task teams,j it is almost irnpossibio to sa3r, but it is n3r guess that,
conflicting recommendations which will require reconci- i.at-i on.
more, they do not necessarily reflect T;.my oun v-ie:,rso
MOMMjIVINJOWLSanitized Copy Approved for Release 2011/01/04: CIA-RDP82TOO271 ROOO1OO12OOO4-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R0001 00120004-6
S-L-C -!Z_L, =T
CRAG 12-58
10 May 1958
overall, the area of agreement between the respective findings., con-
clusions and recommendations of the task teams and those of the Con-
sultants is somewhere between 50% and 60%. In some cases the findings
were validated, but different conclusions or recommendations were
arrived at. In other cases, the findings could not be validated, but
similar conclusions or recommendations were arrived at.
4. The collection of the reports themselves is quite voluminous.
Copies of any or all reports are available to you for retention.
5. Beginning 12 May, I will be meetin;; with the Library Consultants
to get from them their reaction to these findings. On the basis of our
discussion, and subsequent discussions with you., I will, with ED/I
concurrence, effect those changes most likely to improve your central
reference facilities.
Paul A. Borel
Assistant Director
Central Ref erence
Attachment
u~ mmary Task Team Findings
TTR-1 through TTR-16
cc: DD/I
S-1 ?-?C R--E -T
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
3-C .R _L..T
CFI IJ`I'RAL INTELLIGENCE AGENCY
OFFICE OR CEITTIT~ L REFERENCE
ROS'I'ER OF '' ASK TEATI CIIA IRt1EN
Project Short Title
Number
1 In tel lo.f ax
2 Machine Use
3 Coding
4 Selection
5 Acq -. Ops
Fiscal Policy
Publication Procurerient Ted Nordbeck
Circulation
Catalog
Reference
11 Info-center
1.2 i?iini c ar d
13 Reports
1!1 Traini n;
15 Customer Relal,ioris
James Chandler
Wal-ter Moberg
16 Correlative Functions William Tidwell
1.7 Organization
oorc1inator
JA - OCR - 3 December
19 5';
(Deferred)
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S-E-C R-E-T
CENTRAL INTELLIGENCE A GENCY
OFFICE OF CE1 TRAL REFERENCE
9 May 1958
ZMICRAI DUM FOR: Assistant Dyrector, Central Reference
SUBJECT s Final Report on Intellofax - Task Team No. 1
1. Membership
The Task Team consisted of the following members t.
hairman, Biographic Register, OCR
Document Division, OCR
o aphic Register, OCR
Industrial Division, ORR
tzr-
Requirements Staff, Office of Intelligence
Support, DD/P 25X1
-]Machine Division., OCR
ibrary, OCR
2 Plnthod of Task Team Onea?ation
ao All members of the Task Team reviewed the Consultants' Report 25X1
the OCR rebuttals, and other basic supporting documents.
bo The Team as briefed on operations of the intellofax system
anBranch of d in the the
in three sess~hes~~chine~1DivAnalysis
Division, in
Branch of the Library.
c. The Team participated in a joint questionnaire,, prepared
with Task Teams No. 10 and No. il, designed to elicit infor-
mation from consumers regarding their attitudes to and
ntello?axitems i
opinions ustio ire, 12 pertained toethe Of
on this q
g rstem.
d~ An intensive, detailed study of five intellofax test runs
eras made, with the cooperation of the ORR representative on
the Task Team and three of his associates. The runs were
S-E-C-R-E-T
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
l r~ of 'chf. C~?mpc~ts W. G:~:Y,up and Y rer studied
Ir:a.da ~:J.tl? c,l? e he
c,?r_xciu~ly ry the 01WIl an,,~.ly3ts ri: compared -aith their own
e ,: el9_e it :~'~_ ~.C3 o ~ 102:,urnc flt:3 in t:.1f; 3'3.1_G5 Gf 'i.T E; (3 rI's:0 St S
not rct3 c3 ccd by the ns ;ere .si,ed and then studied by
the rep:r~seri'c,s.t.3.ve from the Lociu'ienI'. 1.~ e lnn~ The ORR
ne ysts took into consideration the f e' tor3 of relevamep
coy., le,t:nc;sst use for h~3c~'i}1ec1c:a1 ieseecca prrlers; and
title e pansic no
e0 A caries o x'o and-table discuss.s.C,ns : a.s held between mGmoeZ5
of the Task Team. and groups of :~'eouestars rspresentirg
d1 'f even t types of r esea.'t ch problems and uses of "! L'te .of a Z-
These were supplemented b ,,f a number Coy `t zzd .vI d1. 1 1~1i;er~ ~.e rrt,
of various requesters conducted by rneivi- rs of the Tea-a'
i o lAs z n e:' Lmaut q ~ . Couoosi ue Group ~tia s sot '1p to handle
intello -f. - runs in the .tiF'fere, ~cc; 1ir.n-nc i of the Library,
t onsi 1,1 1~,' of ,ne representative each from Analysis Branch,
riacha?,e D_i.jJj.s en, and the ?r::.bra:c ro `.[l?:i.s Gri.?p; rop,e:sented
oil t-ha, T Team r.unt.ticnad for six weeks. 25X1
The Team held several joint meet .ngs .l:th the Croup to
evaua to end the Group pl cpareda final report
of their experiences and recomnc, da"tiorS for the Team.
(,see -6:ppsn:iix i`2)
as member of both the Compoalte Grcn?p and -the 25X1
r7
,
"ri -r J
li YvCi i?.3 vG'~'S
^"t~ LiS "vit
131tC
'
.
.
3
C
Task ~'~~c m:1 1101.6 SeVc-1t.al {~O_L
,,Th.) h d o ;c d ?'ite11s.~?.r~::;. runns dur n the tenure of the
Groups
by n flc~4~ study made of the processing of nto3lafar runs
to Bete m-r lG ?th e length of time c.sc-nsuirm,d in each of sixteen
major stcp;3 of the y ntel:coL;Kz: p voce&ur the cruses of a.1 r
n ~,ii0a l'G delays ., end the I11a!}1: e in which uho process
J.Lti
111,stened. :tor a Pen_Lod of two c:ceks t,68 r.?equests
could be . .
tJx:e sh :'?~ were attached to all 121t e ofe.'; reque2ts, and
,,r,C al 2116 m c cn2t were recur eod for each sfi.:po The
ormj-t. Li submitted to
1r -~.~>;i es up-Li1'ctr staan. xri,c d and
a from the, l"ana`E.i1ent Staff : ?r further
jthdyo
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S -DI..~ ..~;._;..,~
.711+ )I' _ ieu .1 C: (' .f ? 'i'' 1-. ?.r.~~_,a. 1 ?r -??~ 'rt ?'
''ti l
d.
:, " )C'I1 t::{?,'1'J...'?l.i_,: I?_ a r,'C:'i': ! 1..1 _.n'., 'r'`JCI~.'i_ ~C1:~.
{
, ~.. :JLL:.CL 1`I::. 1]1::,`c.f:, fi., i7? ii (.)ii t;Il: I rr(lt.fnC ~ G1? ....
?
C;i
L v;:; 2'.iOt.l.~? 1"111"Lc n{ i3 'C' is J:.'.i book, ,in(, 1T pa r:'U. V
CyOf'Ci_) on t.bc.: i r i u(-,1 (, i? _i ni Om.. ,'1g; -ni'c)
~I] i?iJ coc'e :i') ll]i.i. 1 _'L` r!Cl OLI` 1)'. !..i..i.:f'.:.i C._.? 1?.
U( (-. ...'ll c'.t~ .02 C'Gl I:1:.'., .1 Ti
`d'oe ~Ir os r Group, s'-aca 1', .= ?iX
to ('(?.i..._cc t on t ho. iI?te
~.I1C a a..r_? the , , .. .. I. _ :~ Ti nE. ,r1.F ru; t ;
i.1L3 .i:3.~,if. e.on t'o 711.5 Chun t ..1 E', t;ho ^noc t ii?!f Or
(;! nl'.=..__., del ii."'d.' i'h0"!=:C he WUTII?:i i nt-rci or
('r1t'PI1 ,n1'ci iLt,.on; o 1] 1.1
_ ?'.. i i ::.i.?a.on rit:C; ^.G:... ?`. 3.1;'i.n'3..., '1+~ i f L ;.L1L bc'. c t
t:U'" S o1?1. red from .T. C'tl nel ?.. J7?
1 o L c I L'`) t;'fif' Z'r'i~ues . L r LI,c,L1 o .'.lt,. t ?:.
J?_ ~'?,L1] ro t: A1:i l::l .: t: t.
i;Ia:i. LIL) C,...',C La Z)LIrtXD5. , wh. t.. r,_'..
o.! t?:1'i._..f_' C'%`~,I]51 %Y] ho WCLl::?. , f C'Ci'l rlr'n c-:t:c,
e object e ~~~ 3 1 - 3 3 J
lof Lill'' ;-_ ca .?
]L' .,.~,. i It?il
1'.111 of l'? i ro o(_'.i . ror .i ?. s u:~
of
1:1J,ji'`'t, Il(I t o '_. !'...I? i?:'it'_ I' ~:^UI ~_lil)?y.:.: ldilo i I) o S-'-'
)l..fi!1?I4 C2 .:. . 'y OL+`: is U:`
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
'szTB~l-4
F3,. Senior document analysts should be, Cull- time reviewers,
with grades of at least GS-12
9. For every subject broken out for specialization in the
.tnalysts Branch careful attention should be paid to a
conti_nus_llg pro :.gar, of training, both fo?'i ai. and on-the-
L,
and contact should be established. with the major
11s:Lng cff:t ces concerned. In addition, senior document
;Analysts should be given more formal training in the
principle* of classification.
MAI
b o A full, alphabetical dictionary subject heading Index,
containing cross-references and scope notes, designed
for use by the average consumer as well as by the-
Composite Group.
c. To Pti,.pplennient the fully annotated ISC, an Authority
File containing ne:r definitions of items to be
considered under a particular !SC number and the ':CSC
number to be used for new concepts. (Like the code
book. th:+ e file should exist in both alphabetical and
numerical order and each should be maintained in
duplicate for ready access of both the Analysis Branch
and the Composite Group . )
d. For use of the Composite Group, an annually revised
card count, code by code, to supplement the code book.
C. Organization of the Analysis Branch
4. The Document Division should begin mediate? y to study
the :Caput flow by subject to determine what subject or
area groupings are possible based upon their present staff.
5, As soon as practicable., area or subject divisions of the
Analysis Branch should be made, and a gradual conversion
to subject grouping should be instituted.
6. The subject or area groups within Analysis Branch should
be relatively large, consisting of about 5 or 6 major
break-dawns, each composed of a number of coders headed
by a senior document analyst.
7. The ratio of coders to reviewers (senicr document
analysts) should be such that all documents can be
reviewed, if necessux Y .
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
T1 /i-5
10. A rotation system should be planned so as to permit each
senior document analyst in charge of a subject or area to
serve with the Composite Group in the Library, in turn.
Furthermore, the senior analyst responsible for a given
subject or area should always be consulted by the Composite
Group, and should whenever possible be present at the
Inter: view with the requester in all cases when his subject
per7:.ains to a particular intellofax run
Document Citation l ~~ 3 I ~{ ?a
13.. The Analysis Branch should place a greater emphasis
upon title expansion. Present .3t: ndarc?s for e.-,mansion
should be defined and then studied by the Document
Division, which should prepare a preliminary rough guide
including provision for at least the following elements in
title expansion.
Place names
Model Numbers
Specific commodities and types
Organizations,, institutions, parties,
Ship Names
12~ After the :ualysis Brunch is finally organized by subject
or urea, the rough preliminary title expansion guide should
be greatly refined within each subject. This should,
primarily, be the duty of the senior document analyst for
each subject.
13
Serious consideration should be given to the training of
some of the coders in citation methods t for example, a
tour with the New York Times Index,
3_tr . :1 w(:sking group of CODL`C should be appointed to discuss
n :hers of tittles and the relationship of subjects, to
documents. This group saould. request that each participating
agency i:mUtruct the field to use more descriptive titles and
to reduces the number of documents with miscellaneous or
11umer'ous subject s
1j.
,i'ter a f i.znr. policy for ;itie:i and title expansion is
established within the Document Division see recommen--
&.ti.ons ,,o. 11 and 12), F.% and CO/C should be requested
direct ; Lhei-r r epos: s sections, LOG headquarters to conform
'L4 -
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
`1TR/l-6
Retii'eme (y 3 - 9q
nt of Cards
17 Y{L?icrofil.ming of retired Intellofax cards should be
edficoritiriued.
I' . Intellofax Service In the Library. C-Pp yS -- S l)
18. {11.1 iritello 'ax requests should be processed within the
Reference Branch by a Ccmposit,e Group consisting of:
a. One senior document analyst from the Document Division,
serving on a monthly rotation basis.
b. Two or more reference librarians, at least one of whom
should be per..x ne:;-ltl y usi.signed to intell'._ofax work,.
c One reps esen tati,~ e of the Machine Division, available
on call in that :l:ivision when needed.
ci.. One clerk from the pe' car once ixa..c:j
j.9, The duties of the Co posite Gzoiro should include
responsibility all inteliofa:x service and for
ga the: i.ng `? nfcrmation from that sew vice to demonstrate
the various uses :..nci probleme of the system. `.h"uis task
entails bhe following:
LL. Exceptw where extremely high priorities or {,crash"
dteadlines are Inrvo:ived, no request should be processed
for :tntell ofex searvice without a personal interview,
or at least a telephone cc;:i.i,, +.iith the
o.~_ iIIc1J
rcoue.?ster. In the caace cf natl-CY,% request . the
consumer t:should b::- asked. to come to the fibre ry; and.,
if this iu not DQ$sible, one of the librarians should
C:o to him.,
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
Tile Composite Gro,u.p shoals. tni.e the initiative in
tr..aintairiing contact with the libraries or other liair3on
unitt3 in outside zq.,encies, and should fully brief
'those responsible for For. F:rcl ing requests to C71..
c . The Group should conduct -:. selective foilow-UP
program desig_led to prow! _,.e continuing iriforrr tion
ati to the puraoces, type:3 of requests, and. results
Of itrteJ..lo:~ax runs (soe rzccmmendittion 2b, above)
d. The Group should keep ull reports of interviews
and such oche::- records of intellofax runs, i;:s are
deemed necessary for continuing study of the
?inteliola;t process, (see l~=:co miendation 2b, above).
20. he specific duties of the various members; of the Composite
Group' should be as
a. 'Die l .brar=ians should mak _a decisions regarding the
:i eusibility of an intello:~ax tape and the relation
of the individual run, to other sources of informaticn.
b The Group's senior librarian should exercise
direction over the procectires involved in nrking runs,
a nd should. r esolvcr all di:?fereri; e , of opinion writhin
the Group. (3 is er" ie should be at least U-12.)
c . '.i:_j member from the Analy :3is Branch should share in
3.scussionG W5 th reauostie .'s and participate in
all 6
xz ocessing requests to ca:rple tion. Final decisions
:involving, code selection : nd arrangement of subject
.
groups should be made by thi c raember.
1, `i;ze reprecsentati?vc from the "achine Division should
e available on cull to attend the Composite Group's
,;,2scions, tnce--n necessary, to advise on g oupirg or
:i the fe ssibi.l t.y of special runs, and to hel.p
T r a time sc~?.'''uL(s .
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
0
S-E- C R-E--T
Try/l.8
The clerk should perform all clerical operations.,
including preparation of correspondence and document
receipts, stamping of documents, etc. She should also
perform, under the supervision of the senior librarian.,
simple screening not involving substantive decisions.
She should keep all records and forms required, and
should provide regular messenger service to and from
the Machine Division.
21. All members of the Composite Group should be responsible for their
intellofax irork to the Chief of the Reference Branch of the Library.
If, at some future time, the senior librarians of the Reference
Branch are assigned to subject specialties, the Composite Group
should then include only one full-time, permanent Library repre-
sentativeo This librarian gill provide the continuity necessary
for the transmission of intello-fax experience as it is gained,
and for the collection of information about the systems The
requester will be brought to the Composite Group by the librarian
in whose subject his problem lies,, and that librarian will then
be an ad hoc member of the Group for that particular intellofax run.
23o Those specific recommendations made in the Report of the Composite
Group to Task Team 5l (see Appendix 2B, Part III) which are outside 25X1
the purview of the task team should be considered by OCR management.
Task Team No. 3.
Attachrnts o
yinal Report Task Team 1
Appendix T'1 - The Consultants' Criticisms of Intellofax
Appendix 1 2A - Outline of Intellofax Service Experiment
J1-2B - Report of Intellofax Composite Group
Appendix #3 - Intel Iofax Runs Made by OR for the Task Team
Appendix #I a The OSI Study on Document Coding
Appendix #5 - Analysis of Actual and Alleged Errors in System
Study of Pratt and. OSI Projects
Appendix ?6 - Customer Opinion
S-E_CWR-E-T
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
.-9
irl (.1`cit( 1:3
2) L} C1'r;l'L:~( 113 La'I: '1T1?(: ;3a'.t i; t ~
j) 1}:2 U;-
1.) .".he S>cuTi-'. d A;'.1 :i;3
5 / r L, r s
! 3 _(. , y U ~ I~.t'U ti ~?:c J;.'"G)'', :Li1:,1 C:c)~ i:~.~';
U) i-: %) L G+?I' ca::'C?S c+re 12:7i1V =?.1].
}`~ J.~'?r it Ci 1fJ L?_CFl'l;G'Li UC7i."~C (:i~11c'. 1!l 'Fd1C' Ci'c;}i,~Y LX 1";O}.:~51
L}1'' S %S'1," i {. S at. d;
a'-i is /1C;ededl f C ~;1? n pro .2' 11 C _ l .:g1=:nc,r'
~. ~' ;
i,.c C17
Sc vtl''c'.t7 l' _. is ar.3:itL,('(
,5 ?~~ t0,' :p i rC .^,C,iipi r -3. .l. )1i, I'D .)I'' 1lili?~_:LE:('.?, CUf?.:' xl':1OTl.a]_
are tlic,. a tr'4'i:!ia 1i:1~Ti?(j 3 .t ."L1.11c:I .
C1.ec.l..f:}~.rE ilCc.:.. "~-O a;:iz; '~I'.2~.:i?:' I7'_I1?Cil_l1'i?~~: c:rif 1ar',CiCi:i t3
w;'( ~'Cti O ).?^aSO S 1 lI?.4l) :?:. the :s L.3.nt- behj nd Illoz , o1' ie_m
.1.(^.r!.,
-0 (% jJ1l;ji }D 0 :'_laJi -u M.,; 7I VB;:1:1.CY;1; of
_t:'. I'r r `'tl( 3'? f~0]_Ili i C th'.~ ?i~~~.~'t1T c+_[C71C: ]"C3
_ t, t e it Z? _'!3 '~_ C 'i.I t%_'1'Ii ,"-_1.1. J:,2~ zJU.,(; L.;~ ~+~ ?I?i:E.:
''.I'i'. 0d, C G 1
(} (::'.Tl!1^ .? 'tlC lli'1Cj'4'3.'"ioo CSI'' r c_ll l ;:(i. 11:C'c?}iClli; O. s'l:.U.C:,,' G:- }lfc=
t r:~ i i:E9 1,01 ,, 3Y'3 %,Il, J11:3t ?):lat
tlt:.1; 5:.., I;?: nJ' Si'a"J .LC L. .l. ti' :! t ,.I,( ^f c;Cl ~C) l"sr )v1
lo-f
c s 1 of ?1 ('1sclis
) C;-Aid LIInc 1;a.c:?tl on
as been ..'ccused on the 1ibj ary cataio,; continuously 3inc:e
tt?e mid nlr_et cr~th century.. In "act;, it :L 3 no exaggeration io
say that throuf;l'_ the ; -!Grs, the catalog l1a:1 :peen a favorite
whipping boy. At 'Limes ?1:'11"13 criticism has . C123come so 'ua_olerit
that one is inc;l:inect t~i tl-i:i'_.ilt U:at -of!lc: r: J_,ht really
prefer no catalog at all to one ao full. of d fects, nl
Tuber fu..r: tin cited alt one of the , for 6jJ'.f coltie , the lack of
?n.foiT?ation P.,-3 to the use to Yrh:.ch the c:~.tr_rnultaneo:usly tsi'hhir the sarrw ovex-all i
nc3e-~d.ng
structure :. The n lyst sy ster!i is,, therefore, a per:,.-,anent and neces-
story part of the g,zneral_ inx:or_m:ation c~c: We may say as a first
generalization that the function of the analyst system is to provide
detailed pattern of ind- :. ng to ansuer a specific research problem
Di s1~ ,~;?c? of ti 1.' s f~ inc: _pl in the t:ant has, - ed to conf'asion
The assumption has been i ade that : a`. the researchers t7ere confi-
dent that they could recover just the materiel they w nted from
the centta--n L sysb ?ills _Lh('e - they would not r .,i ins do their own files.
This was 'ur~~>licit in tho expe.-imen:-,.".. period of inter! of,ax when
i;he vcearL`.hers tiere abstracting anc. coding 'the documents they
Ta;.s12ed to place in the (-'e'ntral, sy;teem. Each was given a code
de iE nation fcr '.-'is o ci input, and tic theory was that he could
hen use the c.-L- 0t,ra". s;,stcrn to recoler' just his own documents
i?nd i30 would rot have to r,.a"intailn his own f]-Le,, However., this
sy;s"t,E:: t coat~:ir:e::t an ina c:rent logical cU.fficiilt.y: each analyst
;r__ahect to ,^:' to i;.cl.t;.n(.c. ,,{3
uh-,-xc 7.0 'a iv a ?5"~s
de '?.::1:'.'S.'? ,')s' ':In.r: i-1 'v ricer
)_"Y' d l Lit? "lr to b:;~:r~j 'v :?:13rP-.,?c ~ ~I:'. dC?(-- i i~"dent L'i:J.():'S F1u2t
i''01.3_+:lic' t3Ct:_s? cst'' " ~.);r F:Z't ),-1 f.';c, :~i:): C ..3i 4:c 1( ?~ i'? us- of ::tt:'Y'?z!f,3-
i7:)''~ L,-,e L X33 Cd)i:'',;:L'?:'.n-t E (; %'Agti;4`t .`u i.? co
A~'.h . ' Cc, ex- ` 'L ~1t::1"~. ,, 'i~ :t' ' ."_?.:i: C'
't;ioc .i :.C~ti 6}~' l.3'~??:+. !,"~J?Y't~....:e'-C Y.l:l f L t, fP 't:(':.CF rr and C:Cw!l:>;:`stkTr:
??:,~11:. J.i: ftU JJ. 3~: L .):1.. tS? t~:tiL p ~ryltJ"e?
`I. v ?!? .~"~ 1. i .'. to G7 7 ~. jJ l1A `- .L' IC%:.?G. . V" 6'..
~~?
or, r , >t
;it.':. i'`~ 7('3 )Jtc)':i l!:_ ti .! t": .," ' ''.J11SCe;. " lr ;\ .vt?~L; '. Yt :a of reporting
c.vcn in tht. ( c r _>r.::;l:'? : t eld : sc--m 1r+c.0!r1io irafo:rmrition IS
c~ tti c:r~ by r~ ~c to 1:f.fr;.ti in tbe3.i a'iri th--rr*fcre
ra Prmd.t)tt?'.ff 3 ?thea `a:tc')n,cal. literature abcut which Cobla s
'u.''?il Cir(:.+':L.7.LL3. .L_(.~i.?I.?_~? c^ ti; .3~?.`a.'.'_ of fh:~ z'Epcx'ci.ng i Gt~..
by nort~?ex;?c::: ?,a;,1;;z :E:) ,t:L wtaotjcts thc;1 n,. e not entirely
e pr; k :,t .y e3i. ct,. rr, r_nd ~c acme o: t?ze pweciaion of t: rrttk
a? lost, o:~' i1:,l :,:nre cLaa?gercua) the tcrmo are uskc,c?
4ri.thou'c, Unr~. ?:Q 3ti;i tt; if tde aze codcc. as if they wt?-e
i d greei: ,;.y, re?r;,:*.~, ;'.-Cj 031 real. cu'bject 6p;:Cia1:i.sta is
likely to ecnr = ti -.a :) of i.hz": track.
b. 1 atu'r'c' of 'f 'k2 ~..o?}.(J,j pct r lCt Ob,ic, .^Is
A13 ed ',c tcc: ~3::~;i :^L't.a dis.usced cb sr~ nti:t cutting
ucc i?t'ar'?e ^t abot:r:. c- or intangible
s"b:?cts Jb ;e,,_ .s . Any document may contai.,
oath t ~s :+l w 9
?,~ i(.tex:. _.. c.,.tr..., j L^ai;Lt~sj?~- c~.~clucsive
r'_i f'j? i ic;ls; (:nn La k3 e^ c ic.h . 'b~eeto and objects, sines
hey cn rlt? ,,. t ..t ~~k::lz ex~ z ;~:.~ :shc12 d i z furen . types of
in r .c.~ticr) ard : etr{ctit"-1 scheme.
Sl-rz'cifN--:: "nr-rj:i ob '_?; ..W. i.?n,k?+: hero to include: Organ-d-
vaticr?s, gx:J ie, jz?c:r c, c1 ~aaic ranee-~e,-~c. , thitss,
E(p::-cif:.c cu:x oditia~; nth two I'a? -.~sr~~ ens, etc. 9/ This ty p<
of reuriew-C, .lc^uC:s; i?~'-t;:Ll tr ts- :^.lficity, detail ami relative
Ftabi1ityr (;.ten cc. rx red *. it, cti)s;;Tact or IzLangiti:` subjects.)
i e objects: ca--a be el. hr e u& mode-lierrs which are riot
c P--se.z of :L:-, pz1 i y ^lcisa, but ciaoses
of cul ord ^r fr?.,a the objects. Due to the
i r?
C- T?td--'r '. ce2 retr ietml is iwr'C'
'?.~Lth -:1 an b c' ~3ulJ LCt.:i T'' .,.'PCs''. . Es'y C ~:Ca. :ark ~ oS~.b~,w
FE: from
It i_tc ;'~::oVn``3`?:C-'r7 tE E, 1b-;.I3 t3@fiIIjtjo f3LS2:C..
an org: kizetl.an
or t^ e,1ii_L? "ii:G 15;t Yr :3
n ~
as a tatter of
.l
yr~,a e2 1 rot of y ilZ 1 iu -.lower to a stxcir'i.c
irr.1.l,.i~' y ). aik11~11 7. ttk1.YT+~:w']C ~" t:C)uld be.
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
I" Tij 1 ..22
L.s.st
this D. rck.Zgh i` .v??.eo ? ''Gr:d ~w3C)p that E: KA2o }C)A,Lc 1~r,~;^vL': e
- or t h'1 .' 1 G"IY' of ?t vm cL t"J l'3 b", t, ,r' :? th3 r t r :.
1. i.i F3 ~.a:. +u$a ".. t?Yc3s~.,,:.e..;C; o .i. R. .,1 ::.?f!'Cr not.. t+...:...:. ~.L (k+. ~~ ... C.4.L,
3m too &? cy'C? 6">i1i3' lp 1,(9'G CcL'ce 'x+i? :' b?'ili:?73a"(r `ems 3i; :er i of
rocC':al ~e :~.L^tto n~a .';t t.T?i'.t n e F -3 . +. der?iy %C:; of : '
?`f ~.al1
f7 . _~:3.%"c3,x:S.?-"ia aver ' ??, ~' rye:-"t". T71a t'y; '~?`'J ~`7'~,? i:..v. ... h :i..., $:.+.v ho i'.,?L"';' ,..
tl;_ fc-1: .. tG.
5...1?.L `/ Y'?i >v lrrJ e .1 ~J .._ tril` Y' (-.' ?.i, .,..L.1 : .i
LL?
:i f.: ^(SC i3. n(.'. ? il:'tYi-Ge:! '.':A1,1
a.ii:eLG_'LiCJ .::LL':L 'l?'a~."V ...5 L::'~i' _
~':;.L'tEL]Ii.i:? ?.Ln f' :L"sC'.:.'2J..~ t'.. ?'1; 1:::%::>`; '!" L i"-i; it f:'i':l'Y?r', Lf
U... ....:i Or L31: ~; ... _r i "T!%i:a:'E T.?iC: (.'1:~t; l,'. .___ .,..?.rS l.ti.: `:~.?ii:: ?:?.r.~..: of
a ?n _ .L ~:.:(.:~..:I. l..l.l .,J lJ ... ..i 1. 71'3.....
.. .t:':E.: l.?(J,.,e:lJ.?'1L::Z :-:c;.ep ( .. _l_.1.ro?!'i'.c: -..oi: , J.' e : ri_'`1. j? :.?h..!,~, ... '~. t. ..(i
u 2. "C'I
AIL ?'~ ,: CIL Ei'y ..':T:; ...ia ~2`i:'.'?.~i2'.is:? 1 ~:iC.:~.'.?.. _~':i~ (~:: :i.>; :'t.T
:'_:i:l .....;e'J F?:L'~1: ..? :,:"_~._. :r.I .:i"1.'1]._.t:.Lcn ._F1 dU .r.'i':nt w a. J' :an .. :L c:'~ :i
J,.
1.1::.:., i.a'in . 1:?;i r j t~ ,:i itl'1 r7ur .. 1f : :.'l'.'.: :!~:_':JL .a. ? :L1 n. tile' iu t::Y:?_.
~~te r.,:,.:,.. ~i:(:1? :L h ?;;~,~, iIo: p e " tr,. ..~I~i;L'1:)_ s;i era
Lies c'. 9 for "u11'.Z-'
:Lt?C':? cr c:7_.... Cclt7.._._''~5 j` ?s
.~:.i.' ._...!,('._.'l..y' lV.~? (J'~ .a_'.. .!_ 1. :a.,i ...:::.J~~ 11 .._..7.~~ o C'.(1 .. C1:i~I :,!.1.o ?.t1:it.t:~.
r
:, (;t_ .: h
co f l ]lt'.. ')~f~:i y . o.t.:' f ?' ~:i) %1; ^.r'S r_ `
1]i ..:ll'?,~.ES .. .. 4"G iii J. C, ): :: E.([_,_..:I'lC:'. ..,;1}L!.r :):., T.:?';'1:' ?...?.. .":r 1r :C'~i 17.:''.:~ier
X47'_. ..IiEC) .LCi_ r ,iL,x i e . ...:`..^C.i:e:r' cai!:
r. r .i: l'.: I:rJ~?.i. it:i. ? !). r.: ~. i +. ..a.: Lli _,.:I i,l'... (~..i(7S' .> _. ;:i .~i )~l~;i:~.7. L) j f.'. ",f.!-P1
thu ::1DS'. 1'i~))o .crl ?;1'.._.. tS}.' .ft:1'i:1'.Drw Di hir _1 ~ ; . r+r1
~'ii 2.):'' 9 '. ii:11. :?iL:iI .:i'-:1''. _ ? On r; :.ht.... .
j' i?
i Yi_Ci'~;
0, J, i.7ea'. ~ju:"U::,.::.3i'?:i
i_: ... ~..~. r? 1 :r : ,.:i .i' :%~'i :1.?i,.l" (._: .;~ :1 .i,....:~1:: ~~. :::" 2:1 i-ICl .... 'L :)eo~1
....'oin
n.
11 .1.:, 1. i I ;r..1_:. 'l?i(^ 1.:~... ECr~~ ;ilCl'a~"i VD 1rJ
L t..,
'.R ?.f .. J_ ~-;?:.L. i..(.)1.?.... .... f. CcLIE.:-,.',,..
j? 1i: x:. '~:.)I --.. rr 1;_.,'?'.:. '!il r.'.1]?t)`I:. .il?'.'. I:. li; .J:.1:;' ~s:.'':;';.'~:11 tea`,
...~, ri '.)'.. _: v! tntr?l~?~kl..!': ..~ f' :; :'?. '..s? f. .:'i.:.: t, Ct t? i7 J ]. ~~; L~.I l._..1 ...
?
'i '!'.:. _,.1.. v.. .': )'~.I ,_)1. ._ .. eti .i:"r r... r' .a' ~~ .?+.)...l it f.J'.
_ . C': L.Lei(. .. i... ,,',7
0
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
a
` ''" V::i l:(_ .J. .S.~..-^~i:ir J_ ' ./,}. r.T'. L.-::1`?,.ia
4r'.
CIL
~;.:./::^.: !S'ii.i'. ._.... ~. ??~ 'i,"' ~?r~i.~. ': '1)."t::'_.t,: ,t.';'_i'~, . J
!i~i.`~ .. ~:!.)il ` %~. ~~'J;:1.-` . ..~ir>,i,? ~.,?~ ~,,,,'~, ^1 ?.G;.. Y'..',? l?_`,i; . 4,`.'E ,.?=~? fJ~ ..~~i:..1'_ii: 'L
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 I
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
L~ tU.f: D;:1
r(:r 1 t?:._ ?Iiscs~re? o:Y C. lyo it lI?lCi: t;}If
TA .le, i]:; _ :~.~? ac ;3.t~t;t_. :~. tiYi~ I'.c i;tl:.c }l,>
?.it.J l:i'O'v~aU- 'LO C., i 4Qrtir_?!J @ 3 ,, C..`C71 tSat`. of Ek">FC~.~?'i3~}ii.
i_,1~ho' t _ie''' 1 1:"~lE _c'tc3 b ~, ~, in3i~1C.11R?.f' rosgo;.ttin :
11tr,Ierou~; z;izor ` .:: ~.
to U-h: ;itlr-tf:;ic ~Lt3~-c 'ctil. zeo T1t.:I it ic.ate [. geu~xii}. 13a-:.1Sf:.etton Mutf1S
(, ~Fli;i?'; 'I11L:12Nl3' L:'. IlC{ rG U: tb: Cit^3'`::.z . ltl ?: ?3 :Qt :e to VIC c' stion 1re-
'i:CT .:.:.t .S' VE:C"t'lt:i1 ; i 217 or .{^.C.1" cent cf the : r1t~
`tl:.Yr'11.1'.E:18 tT}t0 t'>tatti'i t:}l [t the u~e'1 ',L~^ ;~'Yt`t]u C[3tt:i. t.b:1t the
rt' E?tti,ttx' a;:ii1Ci^":!'.LO:'y 'Y' 1: thai'. 3C:t'..of scto ?:ti. Fore F:I~;t t1 .-
t:Ftcl i;) t1_ i't13 _ f1 it. Vat- tL: it I:lit y CT the au31.y c?ts SZho tote, thst .E'sa
uhnn 5 Y'er cent cr the CiC':L?G?-:iii; UL'; v: `:''cA;C:nt tO El 3iVCT1 L.LI'bject, m,
'y_:cz 50 Per Cc':Il? of t`-! y lncnt toc rt:nt3 1'E 'i.0 Show t:0
Oil an :i::'.1?CZ. f.C 3'F]:~ Y tlil~ l L:'i:> Ct (!.l J}1L` ~'' L ^.'` :j OL: 33:t c -, a-st they -were P:S':,'l.Li.: C1
La: tiax:'. S?ii 3f1c1 ,'I''i.Y t 1C ?}' "t`w A,-, out in tb,! r-3=f,'
~ftJi :1'irC'tlt'f'ic:l: i19 'il'-. LC'2 .tl~'i; _l::_>ddro M t!'t 'vi'('s1t11CZIL3 ~Jac)bl_(3n;
.T.Ii j(J1': ?_. itl ~?~?i_ 1'Ct:i t,~~?.i i>.i. s ..n.t7.r:::"?~. ti' 1 El G-:'.Cl: \X[L_.':.:.li:y.' iuL E1C::: `aCE
infl:;ic1'1i'.l.C; :r'. h'.rid `:' :,?..:'y:.:;~5 :l''i~s'ich a: 1~:.> JLC1E C~i~?3it't
i vili.i-23 to e.SIr:r? e. vast E:at'1L?uii;
11j.' ina`~ .:1L31 iL: cider co tie;; t :r1.
?Y:i:h Z_i ?~iCiJtilt?,L C!:' e~'le 'Z3s of 'thQ 1;8!3'; t s i, I.-Ile [ .you - Gia1 CF'.a be '`t?. CZC' a :'.,?tte ;lob i'?Dr au%l1 Ets in the
ri:C:L'tici_l Of ayGi.C. W'1 -1-0 cii' e to Lathe Ifi:t;L3.ia
-.:o l ii:ar e^1? (_.:i~~ }l:'~h7t :n3 in r:s'.~:a"cer depth. The '-ix or icnce
In I. i_:'~'l y Of 'GL' Q~ .x-1:..1 }];'L1%J.~Ci^ Ctrl ;1 CC.1t.L8~ C1'3.A:iZ=.'~C:t 'sT~C! 1 = ~''.v^
r;3r31 t r as t:l':i?Z!_T.eil is 'di' to c.ry
1 poa~.fl 31t OI:. ? F~'1 ~y a. C *t:?.o e Cc:13t. c?l 3yst.e 1S
1:t/.1d E3 :.' t, 03' tL
e > "lie 1. ^C'E.L1ti s'3-:C.Y1. r,L ~O:LI ivii .?i?%. in L)Cl' ?t1C! i2. 1_t`".;: caci'?iol"1)
-,-t th 1 -t Y`,:-- :n i s a e Cf i~-Sor?.at:`.c` ~~t%Li S a
7i:tO:t';T':iE' LT?~?, ;~?C'(i E'.Li'a,l'_^.Cc2[E L'u( t 'GcJ )::'t77 C.E' us FitVr.. a -.L;
,' lyha i1 S;e CG: S'L'::3t1GL1.-
R,'_..:Ct~ ill co i f:
T.3 ~J^i: i?': ].fi' G2:LcC`t:1Q3~ =%..'._"L'?`_ `: re,-(Dam c o. LZ53}, e n-3;1 de--"d-line prob . ?113)
on (' %'':1 0~ `''?Ll^ i'.'1`L."Cl: 1?lLr i 3':~f x 1:11'. O::;a}]'L"nC713 iLl c)hic~i Such
_. "C-Trt>
?r>
C G r3 ,,+1p - pre"
~; ~ C ~.'ir!.8 (:' ?~" :.C 1 . ~ ~ .7;' u,L:.C i -.., j Y T..._ le L r'C. .. .
"- -
r7 cc.c o.i 'Gi1G :Tit':.FIJ.j~E ;O ht C tcJ:tC.i~.('^.`Y I}. tl P :153,~,?Ic of Y.J
.Lt3f.r L`C: _C)TI,
}:' d ,t rc::.tazr. IL:i: s~ to ~l:,ce;fi;`-. ,atl.~' t r3iC)1? ,r~~i~: 1c. Yte e ?ss
' `.(l.C:I :.E1 _ Cl. [ IGw`( dC] - 1lC) 1 ' .:'vC' 1 + y:: ."lC t T1 ~5`D Dom'
ilA
~? ifo:. T 3?].i',7_C.ti -t ?i t, "~ , '. UE- OC 3. 3 as Z'.r.(!}r
rC'."l''r. ?.- t.!1 J c.:: p3 -.- a ( 'i.. cSfl 1,T::.c U .. t(:.:- y 1.(.._ ~>f.C:c [......C _ rv,t.v_`.~S1s ,3:'
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
T3 R/1..21
the task teta;a) 'rill enable OCR to rears a judanent as to the type of
ireo_;raetian retrieval system. apish would best sere the needs of the
agens; y .
Basically, the future of the central system can lie at either
of two extremes or at various points in between. One extreme involves
the r.erinement and improrremen'c of the intellofax system of the past and,
ultimately, the further development of these modifications through
the tidrption of Minicard equipment. This would be another step in the
direction of machine mani-Pulation of coded and clear -text information
as a central and eng eralf ose retrieval system. On the other hand,
the central system could be split by making a distinction between named
Qhjects and the so-called intangible or abstract subjects, as defined
rand described in the General Introduction. Named objects of all types
,could be removed entirely or treated separately, with varying approaches
.up to the full information retrieval capacity of the Registers, and
could be subjected to into u;ivc machine control where necessary, probeblY
through Minicard equipmen'. ? It is for this type of material that OCR
can?provido more service to the requester since, an noted above,
heurS tic value is much lcr er in the proces3 of searching for the named
objects. The subjects rcmnining in the ISO when the name' objects are
separated (either physically c.r lcqically) are the abstract subjects.
They obey Lzhn's dictum that flexibility requires generality in the
input., since they are more subject to change t'rom analyst to analyst
and Iran one period to the nest- As already noted, the abstract sub-
jects present the greatest heuristic value for "browsing." Since they
are the most susceptible to constant change, reir+q related to the
tho: gh~ pattern of the maient, their encoding inust remain general. so
that the system can be used to geuerato new files for new problems in
the future. Since these subjects are more amenable to the general
anprcac'a, the role of the machine in searching my be less important.
Also, this type of subject would probably lend itself far better to
more general treatment than is possible in the present mixed illo,c,
being perhaps best suited to the open catalog or prir..ted bibliography.
T~:o categories of ailed objects--People and industrial plants--
have already been seated from the central system. zoth BR and T.
provide full i_nforriation rftrie'.al and possess the capacity to piece
t~ethe.~ fsagmeTtary infoitiation. Another category which might be
broken out and bandied in a manner similar to that of BR and IR is
that of "organizations ans. institutes." At present., no one holds
prime resrcn3ibility to the whole of CIA for retrieval of information
on thl.s type of named object, -which Is amenable to separate treatment
a^_1 to the collation of C se gtccntu. y information. The need for more
control over place nsmes, essential to the geographic researchers and
to DD/r analysts engaged in "reality" _tudiet:i a'c'{ OTl: If
deenlt- d to be a pob~C IcL u "L C'? ent ?.r':[,r.'?'? i IId C~ t ?l ?ipec i f c location
:J n the document, could. be coded in a m7.,I"!ne-r to t ha - done
In 1 ;, so that )_of atjons and place names could b: searched by machine
:?l1Ci hand sorting. T Fails woi)ld require a '_'ePDratc card in the Machine
D'_Vj.f CFl, but ale card would bear the a came .address .(doc-me?.it nurher'
Tld would ?us d:l.:!'ec-t the searche to the Fare LxY3 ''i;Ll3`~ card as the
,,:resent !iub,jeci; cards. '.Sh-!i , process would lead. to en area file of
much greater depth Yen is now possible but vrould entail added expense
i I? Cod? nt,,, and puL '.hin-; ? X. -this pr :blein is cons' d.ered of Insufficient
iuport .nee to ~?fr.~~'~ ,:arrl; such e. detailed approa::,h, place names could be
Ieft in , e p G>G'i:!'f: sys X21 'e.nd. the burden p_ aced on the citations,
thereby eaki nog, poosih_l.e .a11ictt hand sorting from larger pouu1atj oris o
n either case, the c:l tzj ii ons should i cludee place names wheL'L!Ver
e SJ.b e see he7_c%'.T; . iVL3ICcMind of Y is tl objec+, -moo consider is-
t:icxiel numbers c,-,.-, types, e~ 1'c es n n
p ~, ca,'~.~gory for I~..x'!y 0~R and so~r.G
031 stud:.eis. ~k:heoret'_ tl; ~ .y., ._:i1? c;a`?i.~, agcr~? r r:~ulci r: ~~.so 1
~:a:ll~ be fully broken
Ou'c, but the c~.a.m e tented i -c conclusion vita ted aeove z or place names
pail es here: it might b bet le to handle th" s category within the
cyGt-e-7.: The a1t~ :rvna c.v.a ~~:~?. .,.;1.x~~:,,;-ioe similar to those
..1 t-^d `"or' glace ..__:01er L "c' cod ng of mod l numbers proved to be
des r ~.~.:rl` , it- co .-I.d ?hc ae c ,plichecl In one of two ways: by using _Lve
;tires li: l er ii tai !our CI '_r':S s re ~r'3; lC ti. Y'O ' i._1 s purpose (obtained by
reducine the number of colt mns d 'Voted. to source locator, as suggested
in the r?:'l~orrte cl Teem 3), or by additional card. as
'0:0 p1r''.'F ul'.? E,
C., + As in a"1'v case J'' ace names the entire
bin den would br:.(~.1 aced on 'C 4. C:3. ~. ?'i~?.0 15 _'i, were di c ^OV' T'; d i t
C1
ore Cl'. tail.`-..'
S1 F-11e :01-doe ?! i?I, the C ~; ;^'_T13)'~ to kZ:'.} C1? e 1,o h spec _' '~ :" named
/bj C i,~; `.r d ah. 1 a~ "t:.G'?G uDJ'ect:j, and to 3"~ .. e;e. ie'ta_r ial on both a
speC:'_J.i: has re ultr' _ 7 r1 +.S71'. irab:.litu of the
e i' e =I11
rly one of the p'Ctiems uj.tIL C1pi?~.1:"ti "l fit;^:C ss0 The of
i
Z'et ueSt?:: ' C:L'oblem ee ti i:G _' :r. Z" i y f a
i, Q
the rue the
C) the g _ity of his ... i.es!1 and the of the run appears
to be z' cloy ? i i1 to generality ~:hus :i 1. is possible that
:~ J L.! :LJ':Ly ' 4i.' S1 i~ ~
__::var1:::- ' =; in",rercely pr.or%:l..t. n:Zl
1 can be
d1 .'~b
S t.) llii"i..Z'f1t'?e/Cl in i:h:ich "i. .cl runs [:'
oflJ;c..s }':l.'L'liC!t:: ..flc e3J.y 5a i'f'C?I?1; ~iC ltc'.'n or o;il \t~. un
lC;~l,,A?e , .4-th can 7:.._`?[,'i;.i~'.C"t. ;:,7.I~~e ~''... ~!~'}IE'. rune on j-J,`..'i%:,se ngei cars in
V
'l; ES ^y ^i r
.'?iG'IS.E'i 'd.~:1.'~. ~.~~i3i~ hC'we6x ~ _. cent t r3 C! !r
-e t.-w, I`l cT!.i
04 L.1lG s
S,
-r i1 d '. Y7 .., :L' ..1.:.. a fl;;r.' dC:,~1.r. a.: i+.C~.i .. 4?.?tUI?.i'tl.
!' ;il (2.1 T1O':, j. e ?o-?iVIN`, ? C5 r c: l?:L' to a i_ Zi
f Y3") i7." (Il ,, m? e ,3 C~ i fli.': it C? f\ i Cia 1. )_1
f.;C of the ::.end 1J) _t.11 e&'_Ci i.I on
tht'1 i.....P- of :?I. I-L-a i;n i??.'U.': ~:c ., .'~1 COl'lG'_'=1G'i?
t
' .J ~~ !'?' 7, 1L %"+
nfL:C on
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
4 ^:1 hint 0=0 _ t ZC t-L''.+:+;-/ F _ Elm ~ f1.C.:::'v._1'a?+.a
'} Lt~ J..u1?tU't, " :!t...L`. ..'. S .-(:I. to weal "u C +.: ? , t + C jd
;.: } ....1.. t:. < C v..
1 '! tY
J:
' e i.J.:{010 01- 7 ..,. 'ij }.. '~,..1... L'' Z ?'
1 Lill c ` rf?,:.'.2~ 11y; b
other
hurl, Vot .:,.:iu is geared to
.tt . '
C Q iw c: Cl to {=
' ',(, - C `:14 (t: :'I if ?, 1.1. '1 a 1 ? 'bz" coL....,:im ::It:_.L'1.`. _:.~?. M9?,L (ihi ~..'1)t.7:...< , _,., ?~. 1
a t , I r,~:;`z: Li S3 to 4 f; n.- (~ 7y ,
U P& 1, c7 U=T" .r the _.t:+ '' -.4:~ i:_ 5 1
: ti ., !.'_ i?. 4r `7s
vl:i:L;:Ilsi. UeC.'33 ~ _'Y' Am E. ! ....:iC:`?2.`:.'', i!. 1~1:+'y.L r (:: l:.'?"rI...L~LLr 1'J. ...'.`i:: F'~r~i:: }'i.: e~?Li.4 ~.'.y sl!OL. ll. ha UE e "i ;?vn
(O .:. c~ i~~_: .::1t,::;t; .(:~ :n :tea"=":.'?."Lc?'u~' ~yE;~~;e ; .'t;s ?ya ,
iji__'.. '. ~'. ;fit 1..`_i>' S/~~. :i ~..;~i a i,!+Ll.... .. .:. t.'_?.6~t :"ti: _~'.IC vi.. `i \,~i.: ?. 3 .... ... .:.l..UCLt. ' i..I~~
!',?ii 1 .~."...~~:;i L:.c
1 L -2 2' vitit''.a n), o iP`?ae.iG.. ~ .:.3s: iii y- gg tr "x,57.;3^?
z r ~L l.'.1:.....: ' :.'~.::-iii >. t`Lt" c: fi','1 1:1.~~::tiu f.'?`'rzc: ?
t~ nV.... 14~:./.4~~~ ?1.~ u.. ::, G,~ ,:d T~ ii::., ._ i~ .i 4,. Cs L; ?. i?(_Gi h,i F-
'4,J .Ci :3~?, it .a:.:.~] .f~:v'111 .'-vJ i_(. :~'!: 'LLt L.:.i ~t 7~~L~.?. G_[::':i~~'~..: a :i4.,, ..
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
CC
1$ r,,O`1 In by t-_i
;2UZ jl`:,V e I.tC?1;e lri:.?:+.''G !. '~~- ..j `" ` i .r_' y: C.~r+ sl fr .~
C::?11~' i_~8 :i.t7. .LL;-~E BiC:'il .C~?.:3 U`'~" t, .t`..a.. I
li:llC't~ ~h~ y' uer l.l.:~i1'::iy'{jiL? )i7 it! 3u:.:L?c'I;i:1 of t t2e
3 1: 'u_t z
j x, ;, rv . tlc, bey 0?i.:' e. :pax ~:: t.:1: s can -
cerrn d Ili':.h va:^x.o~
.u i fi?i:: G-
L'a{? CC)]C; :=;"1:" .?Cu[ ??ph:LC 'Fe1 'i , c'r~? j2i',~ :.i.EO :C ?' il. "'~,C l'. $':ntz3'(:: :'~ ~vea
?7; ,l "l-cneuE;.i i?;~~.:~_C c~ r.:._ L.?,: k '7~
book 1l.3.. urr3,,l~bi;ed~.a
C'o'r _T,' 1Cr?G' !,oiiie tali' ?t`,ne Lief-I c'iC..lc ').`,; C;C;.Z "at GZS't. the _-,A J -
b~.c=r : ~ :ii3b;.?':: '1: I[i 'iralr r r'C u:.:i''C.? 5y":3iFTc'. As .I.G'tKg a one
21! ~1 'l. .. 11 U L':rC:Y2f l_~l:yl; ce a C' i,ie aeccJ ':L tno2e who require
t,'h. s e 1,11 3 t~ ...(: c(Z s ~e !'e /^ir s p ? ~ ... -~- rev a .: x y
Bite?. gone . :>ut , as point ed - i-at i-Ei tI?-' : C?ene:CZ~? :+t2ti oduC:tioi:.. the
C'!F.~O_ )L?`I)L5e of fh cc-`-x-zi. sysi;e-.m '.s to ':r'G: aex ?il(::r:: ices
C ]L ) `.':1 L !:~) ions, wh d ie '.4-s saC:on,t',. :'y 1i -P is to n!C;; 1C1.e In
their. C".IY,'. f .:.(??i . That: _ r e3 -
:.r:} _
U- ev a? 7 s :Cil a not -710w the, 1I~~C. to so
it sacr_7f'ices its }73"?_i.n_:::vy pu-_-pose. The of ti-lose
CCiici.iit:t.'g's i':i.. -,tto any sG~i ~ t-iC-1 ~ of _' t1-4e ISC nho'-ald be
,.:.~~.ly t1-4e '?' -,
SCl?~ ie ~ ]_' l Gu' e, t)ui: i.t?~',!. L'iL .-I,; ; us to ili'l.r% shou d be Iv_c'iu.1;?ra
code 6:1'J.: C~. .'C.L:''?.1i2 in e Of OCR.. 0 y' 1.I1 v.~:_3
iJ. bu.l.a)?c_ be P!Ll y2"G ._ .i~. ~'t1Oi ~r1 td,)l-: l Ertl.. of co_L'.
an l_. ?.r?'?-tnrsC; t..ali.. `i:c3 '.e .s_. -'r.~, r7` ;1E tr ?.L.(:':a Gj ?t
L. ``( 1;r p }~ .1~ .~.. ti.~.~ [>1 i rv._~L, a? ii ~1 e- s
T1i.'. -' `':'I u:~ la:'rte 81?.l),je o (~s ~1.C C 1ti IY'(_t.! lam' r jy`,, .(_4. o r r4 7ii.'`;t r '~ .( ''. of V -:LU
e_3`L.l :tn ' :L'Y'L.'.n tlu: basic
!1. ]".:i'~ ()C? beti iee,r th .' tciro ':,,;'ire Ls OS' toi terra .. For uhetic r- asOII:~
I C T'!ust .. ., r. air- a -c se. A iarg, p:oh:1cra to be
~Gi3 G.c.:_ cd is of mzn' ' ?e of the nirb 2 _ct~. .t.b j ecte
C. C' if 11? l,C'_. other Ct-d-if lic:t s coui(_ be :L'etiG. d, ,r. e
Ci'E: "ICI'eloprytet;.t of me-ti and I.-he co3xt-.Z.nuoII-_ in
L..';r):': ~..ab:'.C72S1 7.p O 2.i ee?. c.-s' fl1 ' C 1T. 111j}Oi~ ~'_I)1'C
C C. 1 .( .?Yi - ~Yh _ Cta c o' he
'' a:: ?i,(.i ;? f
co'1c.'
t'l~r . The sac 3,.
fn-.: data U:1 f,y1:-... G....yi,i"1 j:;OiiAt,f; '.1j7
ccirt',1oz r 'l?rJ ;?:',c, tn.e e !IIU ' 1; be 3')i?'7- od
C) a.11 1..._J.
t'isec:y ,1
-;_le by a
t of-)'Is c)j, .i. loo Ll3t.... by 011 e`; ..CJIl C1
'v;,v:- .c::31T1 iil t:; .f 't' ..; i1 :111_.:1 t:,: 'ia?',.:.':"~ ''' Pt!. .(&i.'.
(',.. shoulcyi,-aclut
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S-E-C-R-ENT
TTii/l -32
any additional data necessary to explain exactly what that par-.
ticular code number covers and, if necessary, what it does not
cover. Thi:, si ~,?,, d hci p , crycmv come of the tag ueness which has
caused difficulty in the past. Studies made by the task team, the
results of which appear in :appendix 5, indicate that a high per-
centage of coding errors can be traced directly to confusion among
the coders as to what is properly included under a particular ISC
number. The difficulty encountered in the past in distinguishing
c ::.rd rroduction i ,.n Obvious e:.amnle of this con-
fusion. In any case such as this, the decision as to what to in-
clude under each heading will be arbitrary. with a properly annotated
ISC, however, it should be possible to obtain consistency in the
coding of items under each heading and in the retrieval of each in
accordance with the expressed wishes of the requester. The consumer
should receive from the Composite Group a full explanation of all
subjects which have been considered to fall under either research or
production so that, if necessary, he can modify his request in
accordance with the distinctions already made. The selective follow-
up program of the Composite Group will enable those responsible for
the system to determine whether their distinctions reed to be modi-
fied in certain instances. Following the full annotation of the ISC,
a complete alphabetical dictionary subject heading index should be
compiled for the use of all concerned. This will help the consumer
whn is not familiar with the code structure to locate more rapidly
the items he wants.
An-.already existing tool which the task team feels can be
modified and thus made more useful is the file of standard codes,
::hich is maintained in the Analysis Branch. This file contains code
numbers selected for the retrieval of information on subjects not
included in the ISC (for example, the Suez Canal Crisis). Organizations
are also listed in this file, together with the standard code numbers to be
used in retrieving information on them. A duplicate copy of this Authority
File should be maintained in the Library for ready access by the Composite
Group. It should be kept current at all times, with the addition
of new concepts and reiationships as they occur. The recognition of
possible items for inclusion should be made very promptly and this
selection should henceforth be adhered to by the coders and by the
Composite Croup. It seems likely that specialization among the
coders will add to their competence in ascertaining when an addition
to the Authority rile is required. Finally, a code-by-code card
count, to be used in combination with the ISC, will be of great value
to the Composite Group in selecting code patterns with the requester
and in predicting the number of items on the tape. In the use of
all these tools, once they are developed, there should be continuous
fecdback among the coders, the Composite Group and the consuiers, so
that -cho annotated ISC and the Authority File will reflect the needs
S -L -C -?R -L --T
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
TITTRA-33 '
o?' the -"',,13ency as realistically as possible at any given moment.
2. Reconmendat-? ona
Ft. f fter r evi.si or of the ISC, emphasis should be placed on the
preparation of effective tools for its use. These tools
should include:
i. A code book containing adequate and full cross-
referenc:ing f i t sees" and "see al.sos") . as well as
scope notes explaining the interrelationship of
subjects and defining -their boundaries when these
are not :immediately obvious.
ii 1 full, alphabetical dictionary subject heading
index. containing cross-references and scope notes,
designed for use by the average consumer as well as
by the Composite Group.
I:ii . To supplement the the fully annotated ISC, an
f utborit;j File containing new definitions of items
to be considered under a pa -ocular ISC number and
the ISC.' cumber to be used for new concepts. (Like
the code book, this file should exist in both
alphabetical. and numerical order and eac;.h should
bey maintained in duplicate for ready access of
both the .Ane.lysin Branch and the Composite Group.)
iv. For the use of the Composite Group, an annually
revised yard count, code by code, to supplement the
code boo-z.
ENEMA=
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 13
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
E..T
31
Qr > , , yp certain
controls on rn .ts but not coding them. This in effect ? ayes very few of the
former controls in 9tQ Y>,
Attached Yy; a revised Security Classification Code Guide dot
d 1
~
il
e
:
pr
1955
which supersedes x'evl.sion dated 13 Feb. 1955.
4 Formerly controls such as /9 were pexraittecl in combination with security
classification 1-5 only. Effective this date security controls /0, /7,
fu, and /9 may be combined with 1-3, :-,-F, and J-11.
In some cases of multiple control more than one slash z" !y be used to
indicate such controls.
6. You will note in parax ;raph ~.,- above that controls may b:v combined with 1-3
and tba.,t tr and. 5 are excluded. '..sere is one exception: State Dept. Limited
Official Use dl3. be coded 4/9. In other cases where you find it necessary
25X1
25X1
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
Iq
Next 1 Page(s) In Document Denied
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
APPENDIX 1 3
INTELLOFAX RUNS MADE BY ORR FOR THE TASK TEAM
When the Consultants evaluated the intellofax system, they examined
the coding of documents at the input stage and the coding of runs at the
output stage separately. Their subsequent evaluations were based on the
fact that reference librarians did not code the same run in the same way
and document analysts did not code the same document in the same way.
This task team felt that it should employ better methods of study since
it is impossible to predict from a perusal of coding alone just what effect
the variations in coding will have in the final results of machine runs.
For example, two document analysts might code the same documents differently,
as they did for the consultants. Analyst "A" might use three codes, and
analyst "B" four codes; however, both analysts may have used three identical
codes, the fourth code having possibly been added by analyst "B" for insurance
purposes and perhaps exerting no real effect upon the ability of the intello-
fax system to retrieve the documents.
The task team uecided to investigate the intellofax system from the
other end- that is, to try to learn how many errors occur in the handling of
test runs; and then to examine the coding to find the reason nor these
errors.
Five different subjects were chosen for the test runs through a selec-
tion based upon the fact that:
(1) They were subjects normally handled by intellofax.
(2) There were excellent files on these subjects in ORR; therefore,
the task team would know in advance approximately what the
results of the runs should be.
They intellofax runs on each of these subjects were requested by re-
search analysts in ORR responsible for the subjects. The analysts discussed
the problems with the Composite Group in the Reference Branch of the Library.
The results of the runs were then compared with ORR files for the following
information:
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
TT1'/l APPENDIX 3-2
I The number of documents recovered by the intellofax system which
pertained to the subject desired.
A. The number of documents on the run containing valuable
information.
B. The number of documents on the run containing informa-
tion of only nominal value.
6. The number of documents on the run containing informa-
tion on the correct subject but of no value.
II. The number of documents on the run not on the subject desired.
III. The number of valuable documents in the analyst 9 s files which
should have been recovered by intellofax but were not.
IV. The number of valuable documents in the analyst?s files which were
not recovered by intellofax because they are not coded by intello-
fax - for example FBIS Daily Reports.
V. The number of instances where the subject of the rua was also the
principal subject of the document.
The results of the examinations of these runs are shown on the following
charts.
S -L-~ -?~ -?P -T
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
Next 3 Page(s) In Document Denied
Iq
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S-E-Cad-E T
3. In general there seem to be little difference in the title descriptiOns
of documents over the peri.od o;* time considered here. Some of the 1951+
documents ware well coded, and bore descriptive titles r others wore not ?
Some of the 1957 documents have titles Which are usei'11s others could
stand considerable revisions
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
Results of Intellofax Run on
Production of Russian Mainline
Locomotives July 1 l -_o Date
Total number of documents examined
T'~,'R/1 ,ppendix 3-Y8
Number of documents recovered by the Intellofax run which
were on the subject desired.
A. Number of documents containing valuaabl.e information 153
B. Number of documents containing information of only
nominal value
C. Number of documents which were on correct subject 30
but which were of no value TOTAL 191
II. Number of documents brought out by the Intellofax Run
which were not on the subject desired
III. Number of documents which were in the analyst's files
which should have come out on the run because they
were on the correct subject and were the type of docu-
r'ents norm .l.l:y handled by Intellofax.
IV. Number of documents in analyst's file which were on
the
the subject desired but which did not appear on
run because they are documents not normally inteilofaxed.
FBIS
Aim
Current Digest of the Soviet Press
BBC Summary of 'World Broadca3ts
M rki.n. ; (:enter Briefs
Polish Press Summary
V. Number of documents in which the subject of the run
was also the principal subject of the docuruent .
VI. Number of documents in which the subject of the run
was not the principal subject of the document.
ohm
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S,-:2-, -11 -E T
`1TR/y ..ppendix 3-9
Date
R_ilts of Intellofax Run on Chinese Mainline Locomotive Broduct3,on9 to
Total number of documents examined
Number of documents recovered by the Intellofax
run which were on the subject desired.
A. Number of documents containing valuable information 53
B. Number of documents containing information of only
nominal value. .~~
C. Number of documents which were on correct subject
but tjhich were of no value. 2
TOTAL 57
II. Number of documents brought out by the Intellofnx Run
which were not on the subject desired.
M. Number of documents which were in the analyst's files
which should have come out on the run because they were on
the correct subject and were the type of documents normally
handled by Intellofax.
IV. Number of documents in analyst's file which were on the
lip '"W, subject desired but Vhich did not appear on the run
beccuvc they are documents not normally inteliofa ced.
V. Number of documents in which the subject of the run was
also the principal subject of the document.
v: E. Number of documents in which the subject of the run was
not the principal subject of the document
SwB-C-RmBaT
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
TTR/l Appendix 3-10
1. There is no simple way of locating the documents listed on the
intellofax run. This makes it difficult, on long runs, to check
off the documents against the listing itself to make sure that
all documents have been received.
2. No pane numbers are given for FDD Summaries, etc . For example
(item #163) CL FDD Summary 1/936, 23 May 1956 was listed with
no page reference. This document was composed of 140 pages. It
had 24 pages on railroads of which 22 had to be scanned before the
first (and only) item on locomotive production was found.
3. The FBIS documents often actually contain no more information than
is given on the Intellofax run (Item #170, 171., etc).
4. Air Intelligence Info Reports which consisted of approximately
166 of the 430 items listed practically always contained information
of the type requested. This was probably due to the fact that most
of the reports dealt with single subjects and were short (about 1-3
pages).
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82TOO271 8000100120004-6 -" `"i~
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
Iq
Next 2 Page(s) In Document Denied
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
A11NEX /r1 TO APP1_;I11DDC 1/3
CON5UP1ZIt AW LYS fS OF INTELLOFIt.X RUMS V1ADE FOR THL TASK TEAM
The five ORR test runs were made for the Task Team by four ORR
analysts:
who directed the investigation and also served as a member of e as z
Team. In Appendix /3 the results of the study were compiled by the
Task Team., which made no attempt to interpret the findings,, The follow-,,
ing report, prepared on behalf of the group, includes
a more detailed summary of the findings as well as an inte ?pretation of
the results and thus presents a judgment of the effectiveness of the
Intellofax system from the consumers ? viewpoint.
Introduction: purpose of the pry, jest
This research project was designed to -test the effectiveness
of the Intellofax system as a method of document retrieval and to
identify its strengths and weaknesses o Speci:"ically , the project was
to determine:
a. the degree of pertinency and relative value, of
documents retrieved on specific tape runs;
bo the number and type of documents which were not
re tri_eved by these runs but should have been;
c., the volume and type of material pertinent to the
subject of the runs which is not now .a ntellofaxed;
d0 the consistency of identical runs made at different
times m
2;. Procedure
Five subjects, corresponding to comra%)n types o:t OFdi reports,
were selected for testing the system, T'uo f actor. s Love rned their
selection. The aiialysts who were to carry ou L the research were to
be experienced and thoroughly conversant with their subjects; and
their files were to be well--organized, compre`aensive, and to have
been maintained eonti.nuously for several year: 3. In addition, an attempt
was made to sample. all major areas of the Sino-.Soviet 13'_oc (USSR, China;,
Cz,echoslovel{ia), to vary the ine-span of the test runs (from the past
two to the past e .ght years), and to include both very speciC:ic and
rather general topics. A description of the :,objects and a listing of
the analysts who performed the research is Liven in Attachment A.
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 The testing was conducted in the following manner Each topic
was carefully discussed with the Composite Group in the. fefea-ence
Branch of the library, coded, and two copies of the resulting tape
run obtained. The cards were not screened. All docw-i:nts appearing
on the runs were ordered.
In order to establish the degree of pertinency of the documents
which were then obtained, a series of categories was e,:tabl-fished into
which the documents were divided. It was assumed that the analyst
was screening and analyzin r documents prepay .tory to writing a report
on the subject of the run (see Attachment A),, in the screening stage,
the documents were divided into two general ?;roups : those pertinent
to the subject of the run (i.e. those which had been cc'x?rectly coded)
and those which were completely ex raneous to the sab jcct of the run
(i.e. incorrectly coded or appearing on the run by ,accident)
To deter nmine the relative value of the pertinent documents to the
hypothetical reports, they were further divided into three groups:
a. those which contained data of such value t .at they would
definitely be used in the actual writing of such a report;
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
b0 those which contain :Less important data useful prim rily
as background information;
c. those which contained information of such unin:portan.t or
fragmentary nature that they were of not significant value
(although still properly coded and belonging on the run)
The analysts were well Eruare of the subjective nature of deter-
mining the value of particular documents. TA is was mitigated as much
as possible by exchanging views on how to apply the criteria. Because
there was undoubtedly variation in the application of the criteria,
the data in the attached tables should be considered as illustrative
or suggestive rather than as definitive
In order to establish thoa number and type of documents which
should have appeared on the Iritel.lofax r'cuis but failed to do soy all
documents in the nn.alysts ~ fi:Les were checked against the documents
appearing on the test runs and doct gents of types normally Intellofaxed
but which did not appear on the runs uere identi.?'iecl and listed, (These
lists were turned ovt r to the Chair m n of Task Team One for analysis by
another member of the Team) o Alter listiriti;y the docuniats were evaluated
and divided into two groups- U-iose which ti weld he used directly in a
report on the subject of the run, and those which would be useful. pri..-.
marily as background inf ormat:i.oi:n .
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
IC --IIILI-1-11--T
11T.'ILX f 1 - 3
To identify the volume of material pertinent to the subject
of each run of types which are not now being Intellofa}:ed, all other
useful material in the analyst ? s file was divided into two groups
ones material directly useful to a report; two, materiel useful
primarily as background information.
Finally, in order to test the consistency of runs made at
different times, three of the original runs (nuinbers 1??3 in Attachment
A) were run a second times, using the same Intelligence Subject Codes
and cut-off dates,
3. Results.
25X1
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
iL...I+,. T
TTR./1 - A PPLflDIX 3
iJ JL.;. /r1.s11
A second conclusion which may be drawn from the test runs is that
the coding of documents during the past few years is much improved
over the period before 1951. An examination of the documents drawn
on runs 3 and ). shows a much lower degree of pertinency for documents
coded during 19149.53 than for those coded later. The pertinency of
runs 3 and It was also less than that of runs 1 and 2 because mainline
locomotives are less identifiable from a coding point of view 'F.han
are trucks or passenger cars and some documents containing information
on locomotives other than mainline (industrial or raining types)
inevitably are coded under the mainline category,. A third factor which
reduced the degree of pertinency of runs 3 and 4 is tt_e fact that until
1951, the I. S? C. did not differentiate between mainline and other
types
Related to pertinency is the pr oble?,1 of single-subject versus
multiple-subject documents o Per' Periodically the suggestion is made that
documents be coded so that those with fragmentary info r-na lion on a
given subject could be eliminated from a rune Although an evaluation
of this suggestion was not made a part of this research project, the
test runs showed that some aspect of the subject oi' the run was the
principal subject of two--thd cis of the pertinent dociu-ients which were
retrieved (Attachment 0)0 Again a contrast as apparent between the
more specific and the more general runs. Among the former (test runs
1_1j), the percentage of documents in which some aspect of the subject
of the run was the principal subject of the document ranged from 50
to 86; in test run 5, only 16;'~ of the pertinent documents fell in
this category-.
So far as the relative value of the pertinent documents goes, a
surprisingly snail percentage was considered to be of little value on
test runs 1=-4As Attachment B shows, 6 out of 7 documents retrieved.
on these runs provided either useful background information or informa-
tion of direct value to a report on the subject of the runs, Although
test run 5 illustrated the greater ctiff'icultir of procuring useful.
info rruation on a more gsnera:L topic? even here more than half of the
pertinent documents contained useful infor:na-;:,i_on.
b0 and c., Pertinent documents which were not retrieved or
retrievable by t 1~l toll.c,_~~; _ sys em---
The chief cor::clusion to be dra,:na from F.L Comparison Of
the pertinent material in I/T I files with the pertinent material
retrieved by h of the test runs is that the ? ntello.f ate: system is doing
~- r
a ve :r sat:! s.L actor~v job of rc'tr:ievin, do cumer:ts which are placed in
the system? As lttacnrnent D .shows., the fi_i_le;; in .:f/ih conta ned 125
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
TTh i - Its PENDIX 3
ANNEX //l-5
documents which apparently should have been retrieved by the system but
were not so retrieved (excluding test run L. for to;hich this information
was not available). These amounted to approximately 18 percent of the
pertinent and useful documents on these subjects (test runs plus addi?-
tio nal documents in 1/TH).
However, about '0 percent of pertinent documents in 11TH files
on the subjects of the four test runs were of types which are not now
being Intellof.axed, and of these, three out of five contained informa-
tion which would be used directly in a report. Data on pertinent
documents in I/TH files of types which are not Intellof.axed are sure nar--
ized in Attachment D; the documents are broken down by type in Attach?
ment E. As the latter shows, FDD and FI3IS publications comprised more
than three out of five useful documents of types not being Intellofaxed?
d. Consistency of runs.
Test runs 1--3 were re-run with the same code patterns L. to n
weeks after the original runs In the case of run 3 thc3 terms of
reference were sliL}:.tl-y different, due to an err and then. afore a
different number of documents appeared on each czar. in the case of
runs 1 and 2 both versions were identical frith the exception of docu-
rents added after the original. cut-off date.
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
Attachment A. List of research analysts arcd descripf.:i.on of topics
Description of topics
Topics 1 borough 4 were selected as typical exar le r
tonics dealing with readily i(flen-bifiable ' '
Topic 5, a much more general subject composed of less easily identi-.
fiable and less specific elements, was selected xor purposes of contrrast~
25X1
25X1
25X1
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
Next 3 Page(s) In Document Denied
Iq
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
go.
S _E -C -ft I-E, .-If
APPENDIX #4
The Oil Study on Document Coding
In 1956 Dr. Malcolm Pratt of OSl conducted a study of the intelligence
documents received in one division of OSI (see attachment 1/1). He had the
subject interests of FSA/OSi reduced to code numbers in the ISC. These
codes, representing the requirements of OSI, were compared with the codes
assigned to the documents by the Analysis Branch, Document Division, OCR.
The conclusion of his study was that 40% of the documents reviewed (which
were pertinent to OS1 needs) could not be recovered by means of an intelm
lofax run because the proper code numbers had not been assigned to the
documents.
During the two years since this study was made some individuals have
accepted it as proof of the unreliability of the intellofax system while
others have questioned the reliability of the study itself. However,
until this time no one has ever examined it to determine the validity of
its conclusions.
Fortunately, Dr. Pratt had kept the material used in his study.
D The task team was able to examine the codes used (which OSI had described
as equivalent to their reading requirements) and were also able to scan
the documents which OSI claimed to have been coded incorrectly.
To evaluate the study the task team took a sampling of 43 documents
from the group that OSI maintained were essential to their operations)
but had bean coded incorrectly. The coding on these documents was then
evaluated by the Analysis Branch, Document Division and by members of
the task team to determine whether it was correct. A detailed break
down of th:a number and kinds of errars found is included in Appendix #5,
below.
The conclusion of our examination was that the number of documents
coded incorrectly in the Document Division was about half of that esti-
mated by Dr. Prate, He stated that 31% of the documents essential to
061 were coded incorrectly,2 while our sampling produced a figure closer
to 15%. There are several reasons for the variation in our findings.
l None of the documents regarded by OS]]: as marginal to their interests
were exarni.ned0
2 The f igw?e of 31% does not inclu!'e the items of marginal interest to
OSI. The 40`, error mentioned in. Dr. Pratt ? s report includes these
marginal items
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
TTR/I APPENDIX 4 2
First, the codes submitted by the OSl branches as representing
their dissemination requirements were incomplete. One example, one
important code for nuclear research was missing from the list submitted
by OSIL A large group of documents which had not been assigned the codes
specified by OSI could have been recovered by an intellofax run made by
a person who understood the use of all of the codes in the book.
Second, several of the documents did contain the proper codes but
the clerk in OSI who conducted the experiment had failed to notice them.
Therefore, whereas Dr. Pratt was correct in assuming that errors
occur in the coding of documents for the intellofax system,, the t.,tal
percentage of error shooun by his study is not as great as his report
indicates. For a more detailed discussion of the errors and for sugges-
tions as to what c.m be done to reduce the margin of error, see
Appendix #5, below
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
SG;C_C4RyE_T
AT TAC'H EN]I' 1/1. TO APPS 1 D: 1Zt
13 November 1957
1o The attached Figures 1-3 relate graphical3y, types of
CL'L and Air Force intelligence reports with analyst interest
as these were routed by OCR to the Physics and Mathematics
Branch and the Chemistry Brancb/FSD/OSI during a three-week
period in the spring of 19560
20 The study was undertaken to determine whether more
effective dissemination might be achieved on the basis of code
assignments under the Intelligence Subject Code (ISC) in place
of the subject reading guides then in use. The conclusion was
emphatically "no9S9 more serious., however, was the finding that
of the reports judged by the analyst to contain information of
which he must be aware, some forty percent would not be retrieve
able through the coding and machine control as then maintained
by OCR.
3. The test wt conducted by asking each Branch to desig-
nate all of the codes in the ISC which covered the analytical
function of the Branch. As the mail was receivedD the FSD
screening clerk, who worked in close association with the a
alysts, separated the reports into two Piles; one c!iich she
thought the analysts would wish to receive, and one which she
didno.t. The second pile would norrially not have been sent to
the Branch, but for the test period it was, A second clerk
attached to each report a 5 x 8 card of different colors de-
pending upon whether the report carried one of more of the copies
specified by the Branch and whether it would or would not nor-
mally be routed to the Branch. The analyst was asked to detach
this card and check one of the four statements thereon, which
were in effect the four conditions shoe n in the legend on each
Figure., cards wr, return6d': to, a ce ,tra1, point for the
analysis .Mich forms the basis of these chartso
4- The test was confined to CIA and !Air Force reports
because only these were coded before dissemination by the OCR
reading panel-.
Fundamental Science Divii lion (FSD) was then known
50 The
as the Fund ntal. Science Area (FSA) 0
6. Liter ally, the notation under columns 3 and 4 should
read "'not normally routed to the Branches"
7,, The forty percent free, mentioned om!lier,, of reports
containing essential information and yet not later rotrie le
S:_CcR'E T
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
TTR/1 APPENDIX 4 ATTACHMENT 1-2
by machine searching is the ratio of the area in red to the
larger one outlined in red,
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S-E-C-R-E-T
APMIM LR #5
Anal s of Actual. ands, Errors in System Sip of Pratt
Z'ro acts,
The charge in the Consultants, report that the system is unn
reUeble is meaningless- it merely implies that the system is im_
perfect without giving any degree of imperfection.
TWO questions must be anmered:
In, What is the margin of error in the system d 5,'.' - 10% m50%?
20 If the margin of error is greater than can be tolerated by
the users., can it be reduced by a) changes in the esteem
itself- and b) new controls and checks imposed on the aystem?
To answer these questions the Task Team used three met nods of
investigations
Io It made test machine runs on subjects where the results
of the runs were already known and could be checked against
ORR researcheral files.
2. It reviewed a study made by Dr. Pratt of OSI several years
ago in which he claimed that the coding of documents is
r rry unreliable.
3. It interviewed analysts from various Agency offices who
had participated in the questionnaire to ]learn how efc
fective the system is for their particular needs,
What is thaoosen`, margin of arx?or?
An examination of a sample group of documents from Dro Pratt ? s
study indicated that about 15% of the documents were attributable
to system errors. In one test run (ORli s #231),9 the margin of
error was also 15 %a
in 1956 Dr,, Pratt conducted a study to determine whether it
would be possible to route documents in OSI according to the codes
assigned to the documents by OCR. Because he found the codes un-
suitable for dissemination purposes, he concluded that a large per
centa,ge of the coding was Incorrect, A i.nre complete keview of
the Pratt study is contained in Appendix 4 of this report.
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
II
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
TTR/1 APPIJEiLIX 5-2-
SSE-C-,RAE, -T
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
031 (Pratt) Study - Errors
Sy?.Jtern Errors 21
Coder did not understand document
2
Coder missed an important concept
3
. use ~ enera7. codes rather than specific; therein
ore missed specific institutes (carelessnessg lazi-
nessg misunderstanding on extent of coding expected)
3
Confusion between researchg, development and Pro-
duction (600 vs 700 Section of ISO)
7
6
ropriate codes
Coder omitted ap
m ry
p
listed by OSI
Open publ ?.cation - not coded in system
Missed by 051 clerk
Non- rsJteiri Errox?s tY 22
Content of document: did not warrant use of numbers
Errors in I`iach Ruu
Syst~'m errors 33 (3.405% of total rte)
Coder did not understand docmin -Mt
f-od?id not include code for trucks
co as u:3c+d are on docurr~'V:;) lainess9
carelessne=ss :tack of understhndingo
Coder.Ts failure to code mend, or mach-Ito error
(no taa:j of dotern iang)
Machine error (correct code or : ccd.e hsheet)
otal.
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S -E -C -R -E -T
No prefix action code applicable = 5
No specific mention of trucks, per se, in document 7
Document discussed estimated production, which was
not requested in run (correctly coded 8-11/) = 2
Documents covered other aspects:
Military use of trucks =2
Research and development (correctly
coded 662.113). Composite Group may have been
at fault by not inquiring whether prototypes were
to be included. Researcher obviously expected
these, as he filed them under specifications, 2-5/ ='.l
Duplicates listed by researchers 2
Documents not included in system
(Joint Weekas, Cables) 3
Total = 32
System Errors 32
How to correct the error outlined above - The terms and the area of
coverage of each code must be carefully and clearly defined so that
d} Another large area of confusion In semantics between research
analysts and document analysts (coders) is the definition of "exl.>orts".
Some ORR analysts expect to find offers to sell a product with com-
pleted transactions, whereas the Document Divisioe, coders have the
two concepts under separate codes.
WO&ON Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
'QTR/1 APPENDIX 5-4
the research analysts (users) will know what to expect from the codes,
and the document analysts (coders) will know exactly how to use the
codes The Librarian must continue (and this will be improved by
participation of a DD senior analyst in the Composite Group) to ad-
vise the researcher of the meaning of terms.
This task force or any other could never reveal all of the ex-
amples of errors (or what researchers consider to be errors) in the
system arising from semantic differences. The best way to uncover
this "unreliability" is to have the composite Group make frequent
follow-ups on the results of machine runs. When the Group discovers
that an analyst did not obtain certain documents he felt he should
have received, OCR/DD should either make a decision to code this
category of material under the particular code desired, or put an
indication in the subject authority file (which should be establi.shn-d)
that this type of material is covered by another code. The decisions
so made, as well as those which already have been made, must be en-
forced by constant review of the work of the document analysts by
the senior coders.
A much smaller group of coding errors(about 8% found by the
Task Force) were caused by the fact that the document analyst who
ed them evidently did not understand the subject discussed in
e document. The proposed plan to break up the Analysis Branch
by subject will help eliminate this type of error because each
analyst will have a smaller number of subjects to handle and can
develop some specialized knowledge in these subjects. Again, re-
view by a senior document analyst who is highly specialized in a
specific field will increase the reliability of the coding.
Of the total number of errors in coding, one third were at-
tributable to prior careless or inadequate coding. One example of
this was a document describing several scientific research institutes.
The document was coded under one general number for research insti-
tutes, instead of under the specific codes for each institute. Re-
view by the senior coders seems to be the only solution to this prob-
lem.
--N
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82TOO271 R000100120004-6
0160
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S-L -C-R-E-T
T`m/i Eii?'r{; DIX 5-5
questions are put to the intellofrax system, In the example cited
above, it is just as likely that the agency will need to know ghat
equipment is contained in that particular institute as it is l kely
tW ileea ow many
The present system of placing the senior coders
1,110 erence Branch on rotation assignments should help to
bring the needs of the research analysts and their methods of con-
ducting research to the attention of the Analysis Branch., OCR/DDO
Dividing the Analysis Branch by subject should help also, because
it will bring the coders in contact with their counterparts who are
coding research on the same subjects-elsewhereo
The errors in the system thus far discussed have been, for the
most part., those stem iing from the original coding of the docu--Wntsa
althoiigh rsention has been made of the tafficulties witch arise when
the system does not understand just what the analyst expects from
his hite3lofax ran. However9 in the course of interviewing analysts
from various parts of the agency the Task Team uncovered a machine
run in which the coding of the request for recovery of documents was
inadequate o It is impossible to tell now just how many of the errors
in the system are caused by inadequate coding in the recovery processo
who made a daily review of the coding done in t e--
Reference Branch after the runs were made, estimated that there w
a 5% error on the part of the Librarians due to mi.scoding., omission
of coding., overcoding and misuse of area and action codero The member
of the Task Team 'ho participated in the Composite Group experiment
states 'chat the Groupies procedure has reduced much of the error from
this source-not only because the judgment of two people better than
that of ones but because the representative from the Analysis Branch
is present at all tins instead of arriving at the end of the day
to offer assistance0
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S -E-C -R -E-T
APPENDIX 6
CUSTOMER OPINION
Three methods were employed in exploring customers' opinions and obtain-
ing their suggestions..
1. Questionnaires were returned by 429 individuals representing all
major components except those under the Deputy Director (Support).
Tabulated results of the questionnaire are given at the end of this
Appendix. However, statistics from the questionnaire have limited
value because of the uneveness of the volume of responses from dif-
ferent offices. ORR personnel returned 64.4 per cent of all ques-
tionnaires received. While ORR is the biggest single customer, this
percentage is out of proportion. Of more value were the individual
comments, some quite detailed, made by many persons filling in the
questionnaire.
2. Seven round-table discussions were held with small groups of customers-
four with ORR personnel and three with CS personnel. The results of
these are reflected in the body of the report, and also in the individ-
ual comments summarized below. In addition, an OSI employee who had
made a study pertinent to the task team's area of responsibility some
time ago Was interviewed, his study reexamined, and its implications
further explored (see Appendix #4 and #5).
3. Two members of the task team were from using components, again ORR
and CE, and were themselves either substantial customers or closely
associated with substantial customers. This contributed both a
customer viewpoint within the task team itself and the benefit of
much informal discussion within those components.
Intellofax service is essentially a relationship between the service it-
self and the requester or user. Hence, the customer is an essential part of
the process and observations concerning the characteristics of the customer are
pertinent to the problem. Although OCR cannot control the natures of its cus-
tomers, an understanding of them can add to the effectiveness of customer rela-
tions and customer education.
While constructive suggestions for needed improvements in the Intellofax
system came from many analysts, the more intemperate reactions were concen-
trated among persons who exhibited one or more of the following characteristics:
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
`3-E??C -R-1 T
TTRA APPENDIX 6-2
A nostalgia for the simple life of the days before the invention of
the wheel, the industrial. revolution, or the electronic age. These
individuals identify machine methods of handling documents as the
cause rather than one result of the increasing complexity and bulk
of human knowledge and information. Their ideal image is of a golden
oak card file from which they could identify, with their own hands
and eyes, every piece of intelligence information and every report
and. monograph that has flowed into the Agency since its birth. Micro-
films and sprocket-feed accordion-folded paper are alien symbols and
threats to security. These people merely exaggerate a feeling shared
by most of us, and many of them are quite frank in acknowledging and
discussing their own discomfort in adjusting to the machine in re-
search techniques.
2. A rank-happy di.estasb for getting their toes wet in the same pool as
the humble hevers and drawers. If told that many hours can be saved
in their particular search If they go to the library and screen the
cards, they either send a clerical subordinate to perform such screen-
ing or reject the suggestion altogether and later complain of the
length and lack of specificity of the machine run. The opposite side
of this same coin is illustrated by junior and intermediate personnel
who carry out parts of research assignments under seniors, and who
complain of difficulties resulting from ? inadequate briefing on what
the senior is actually after. This withciraawn attitude has particular
applicability to the passages in the body of the task team report
dealing with the heuristic value of searching.
3. An inner-circle complex; under the disguise of security-consciousness.
Much less prevalent that it was six or even three years ago, this
attitude occurs most frequently in the CS components. It is an
ext;^eme form of the classic conflict between security and efficiency.
In order to "bbu y" the "operational interest," these individuals
purposely disgu ae the request and refuse to explain it sufficiently
to enable the librarian to use judgement and initiative in fulfilling
it.
At the opposiLi Opole , the best informed user reactions, whether critical
or faavor?able. recognized the intellof'aax as one research tool among many, its
present drawbacks including both inherent limitations and correctible weaknesses.
E'vena when present oo; tcomings of the system were emphasized by these customers,
they expressed the belief that no other imaginable system could rival the
intellofax In handling so much and so multifarious material.
>n2> -cthe rUstomer__!-roks For
tirtz r r u~: tcN12e rn neck from the iratel.l.ofax falls into two main groupings :
i. Supplementary information, in eases where personal files have already
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S -E-C -R -E-T
been built up, where other tools are used earlier or simultaneously,
or when the need has been anticipated and partially met.
2. Complete coverage, in cases where the custor.,er hopes the intellofax
will provide all the information he requires.
Several analysts from ORR stated in the questionnaire that they maintained
personal. files; most of these individuals added that they used the intellofax
for "insurance" purposes. Personal files were :maintained to provide an "ade-
quate working background" and to answer frequent "ad hoc questions." One
analyst whose files answer most of the questions which arise had utilized the
intellofax system to check the adequacy of his files and to do research on
points not covered in his files. This individual did not find it worthwhile
to use the intellofax for checking his files because of the time and guesswork
involved in selecting documents. In one of' the round-table discussions, an
analyst from ORR's !Analysis Division stated that he receives a tremendous
amount of material through dissemination channels, much of which he incorporates
in his own files. He uses the intellofax for insurance, receiving a high pro-
portion of irrelevant material from it. Another analyst stated the system
rarely contributed any additional information on a subject which he had been
following for some time, but that he used it when undertaking a project on an
unfamiliar subject. Another stated that he generally used the intellofax at
,.,,,the beginning of a. research effort and. found it gave a rapid and generally
,good listing of major documents. An. OCI analyst stated he had used the system
once or twice for the preparation of handbooks, but later built up his own
files for this activity.
A CS :researcher stated that he used the intellofax and published
rraterials obtained through other retrieval systems as complementary tools on
s. new research project. even with materials in the intellofax system, he did
not expect fragmentary references buried in other subject matter to be re-
trievable through the normal utilization of codes. When he had extreme need
for fragmentary references he used various approaches, such as screening docu-
inents from a particular source of specific areas. Use of such techniques
required vide familiarity with the entire machinery of intell.i.gence production
and was v eI y wasteful of time. It was Justified in instances where inf or:cation
was scarce and fragmentary data badly needed. Another CS researcher stated he
used the irztei.lofax at the start of a research assignment by going to the
refercn ce librarians and erplaiuing his need as clearly and specifically as
possible, then leaving the selection of codes to them. A typical. resulting,
tape ran to 400 items, of which he ordered 300 after screening the tape, and
only a m & T3ority of the 300 documents failed to contribute to his subject.
He ctrated that this percentage of hits was higher than can usually be obtained
fr. orr. the use of conventional cord catalogues.
Mrt.ny users described various needs which the intelLofax t3ystem L~s it now
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
8 -E--C -R-E-T
X /l :,PF ix 6-4
:zunct;icor::, eithe :'ails to meet or fails to meet adequately. These included:
-o
More d tEriled information on cards. Several. analysts suggested a
.one detailed description. of the document on the card, but few speci-
fied the sort of additional details desired. One analyst suggested
a brief summary of the document. Two individuals suggested some key
to distir.G;uish between brief or casual references to a subject on the
one hand and detailed information on the other. Another suggested a
notation as to the exact geographical area covered by the report.
Other sur;gestions for additional data on cards included lists of sub-
topics when the report was so organized, and brief digests of the
report. It is noteworthy that no suggestion on this matter of more
informative cards addressed itself to the problem of fragmentary and
incidental tid-bits on subjects other than the primary topic of the
report.
Furt:Y:ar subdivision in the code. Many brief comments were pertinent
to t:,e problem of tape runs poorly responsive to the requester's
need.. These comments illustrate the range and variety of special
ir.rtc?rests that seek satisfaction through the intellofax. Analysts
in -.,he military field, at one of the round-tables, stated that the
cac'e treatment of military subjects was unsatisfactory in that it was
br:,3cd on US military organization, which in many respects aid not
a; ply to the Soviet and Satellite armed forces, which were of primary
i tcrest, This same group also stated, however, that virtually
entente-by-sentence coding would be required to satisfy their needs
because so much of the material they sought was so scarce and frag-
mentary. / CS researcher proposed adjustment of the coding system so
that it would be possible to retrieve by source or category of collec-
tor. Related to the suggestion of the military analysts above, some
of the CS/CI researchers expressed a need of extremely fine subdivision
in the code and multiple coding of the individual document in the area
of International Corrmiunism. A great many comments had to do with
missed documents when runs were made on specific pin.-pointed code
designations. The difficulty was in failure to catch fragmentary
references, and any solution would require a great extreme of multiple
coding. Two specific complaints were that processing; and agriculture
were not distinguished under the heading of "meat," and airfield runway
construction. was riot separable under airfield construction.
urt'ne.r " o~ r ?ki c subdivisiorrs . A special and extreme instance of
the preceding item was expressed by analysts whose work lent itself
to a geographic breakdown. Several ORR analysts who treat a variety
of subjects within a very limited geographic area said they needed to
be able to obtain runs on specific areas within a given country. They
c:1o suggested that citations include the specific geographic sub-unit
as well as subjects covered. One analyst expressed particular. interest
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S --S--C -.R -E -T
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
in a further breakdown of the Soviet Baltic republics; another
I'::":ntionecl the need for further geographic breakdown of the Western
European countries; still another recalled difficulty in retrieving
Matelrials on the East-West German boundary. Ocean as well as land
geo:?aphy brought out considerable comment. One analyst reported
good results on an Icelandic territorial sea problem, while two others
told of poor results, one having to do with territorial waters, the
other with the Gulf of Aqaba. An hour's conference with the Reference
Librarian failed to produce a means of retrieving the material sought
in the latter case, and the run made turned up only two or three docu-
ments as contrasted to a dozen or more in the analyst's files. One
suggestion was for a "general description" code designation to be
applied to descriptive material concentrated on a small area, but such
ttterials would be only a small part of that sought under the geographic
approach. SR Division of CS is very largely concerned with "reality"
studies. compassing the physical enviornment and daily experience of
the indigines in very limited areas. Such studies are built out of
extreme and precise detail. Such details crop up in all sorts of
aocumerrts, on all kinds of subjects, and in extremely fragmentary bits.
One such study was concerned with a few miles of territory on either
silo of the border between the USSR and a contiguous country. There
was no limit to the type of information pertinent to the study so
long as it applied to this pin-point of territory on the map. Yet
there was no code run smaller than the entire USSR province and the
entire adjacent country. For the code to meet this problem, however,
Yoould require a geographic coding operation of magnitude comparable
-to the entire subject coding operation.
4. Weeding ca ability. This was one of the most controversial issues
chat arose. While there were many references to documents that were
.s:laaply of worthless qual ty, documents that made mention of a subject
out; contained no useful information on that subject, and other ore
that was of so low grade as to be uneconomic to refine, there was an
almost unanimous shying away from any suggestion of weeding, either
(luring the original coding or by Library personnel during the process
of retrs.eval. One analyst who complained that many documents tuned
up in runs that no more than mentioned the subject of the run went
arguing :q(a` us t any wee rag process, one analyst ccmmented he doubted
"that the br=ightest non-specialist could ascertair. tai .t might or
right not he pertinent to a given subject." Another noted that the
analyst hinnsclf, fsnilic r with the material ltj his field, also becomes
i'rnniliar with the relative worth of various types of documents and can
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82TOO271 ROOO1OO12OOO4-6
S-E-C--R -i-;-T
TTIh/1 Ap-mmix 6.-6
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82TOO271 R0001 00120004-6
Use this criterion in his own weeding and selection from overlong
'tapes. Still another referred -this problem back to that of the
paucity-of information on the cards, citing "inadequacy of titles"
as a barrier to weeding out unwanted documents. There was general
agreement at one of the round-tables that the analysts would rather
have no screening done by the librarians for fear of missing anything
that might possibly be relevant.. A group in CS/CI endorsed the thesis
that researchers thoroughly familiar with their fields can scan long
tapes very rapidly, the process serving as a reminder of documents
airead-v seen in the continuing review of disseminated materials
rather tYn as an excursion into virgin territory.
5. Faster processing. There were recurrent references to delays that
appeared longer than necessary in obtaining tapes and documents.
However, some of these instances were recollected from a considerable
time ago, and even during the period of inquiry by the Task Team
various improvements were noted. One researcher who complained of
the amount of clerical work required in submitting requests for
documents based on an intellofax tape called back the next day to
report that this process had been greatly simplified since he had
last used it.
Replies to the questionnaire were received from the following components:
Humber of
Returns
Percent of
Total Returns
ORR
277
614.6
CCl
90
21.0
OSI
29
6.7
ONE
15
:3.5
Cs
8
1.9
OBI
7
.:~
00 3 o'T
].. r. e i_ c~~etzc~yf of use. Question 38 referred to the number of times per
rk-year the analyst uses the library to request an i:ntellofax run,.
Number
Percent of Total
e,.
Do not use
160
37-P
1 0
2 times
155
36.1
c .
? t o .." times
`37
20.2
~~.
6 to i0 times
22
5.11
11 or more tines
6
i_.'+
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
Do not_ u;;
or 27 _5 or I or
110
4
~
c~r :3 , r or
3_5 _ me
o
or 2 8 or 28 ~.s or
on 4 or l~+
S -Z -C
Trr /.+_ PPE1Wt 6-7
L URR
C&
ONIE
11 or ].'
IL
_
2 or-25 %
2
_~
2
2or2
1 or 12"I
2*
iesL3or,~a
or 10
2or25
* One of these Individuals circled this answer because she anticipates using
the system this often in the future. The other Individual. a research
assistant, nu kcs runs for a group of analysts.
2. De,7ree of satisfaction:
k4uestiota is.6. On the basis of your past experience, please
indicate whether the service [to request Intellolax card or tape
runs ] is generally:
~. Unsatisfactory (31)
b. Mostly satisfactory, but not. entirely so (32)
c. Satisfactory (140)
io More than satisfactory (82)
Do not use the service (].)6)
Thoo results of this question as
sin'-A.ar questions dealing with the
shod the following pattern:
_ b it ;~i?.g~a~`?_1;I ...... _...~....._.._
To request the preparation of a
__?_.ePCI_e_d irt..,your *~~o?k
work
To obtain directly specific facts
To cquest Inteilofax card or tape
To obtain a nformatiori on where to got
To obtain directly background inf or-
r ., it ion needed. in your work but out..
c.l_d n, e: ?i pra.si.Tl.g
18
7~" + 4 i'r 7"1 dt T4[ 1 ) e F;ci,~. fio,I 'k-- with 'ii'e7' ' fe'Gr
w F Y 4 r i! G F " i~; Dy ,t h CIA Lib. ar tr and . f3,'`."e L7'1 r ere' ora
~ `?''' " f t--. ;1f. 1 ^,,?x ,., , iCri:?T1 " *; ' .1.L?, 11*Rlt .d..J:. .L. .. tP?1y 3,000 se.r:ial
titles half r~':r 'bese a e Or....F:n:t,:a.;" u;at',er.'ialsn Most; of
t?h . 1.'!' A.11/~.ng . ,1. .I is ~.g; .Lla^'i - in he USSR, C,':1t Eu':,peen Satellite tiro
A* .+.",'-sereal,).. C ri., e .(.1'i,~..Yr, 25X1
ma ntaine J. in - i. Satellite :;ems a.ais are selectively
m.ain:tained.,, Thl,-F :i kept fc'>r the same basis
reason:: are hou.i .e t f re we
` e Re lef.;'f.l 'rt?1~^,tJ'i,? }~ ^1 i n um, t.10 04.i`..: '~.^~J i,,?'?J{/I+.} ~~~- :~).t~.e o; u o n
the gc'1-a'# me j^ r V Jr ~IFh ^ ~!~ f'' yh`)V 3 an n~l i= t i i ':' P_t ' A ia.{ iS L.:? o
-;, .ti - }IIFl r it]" J
A 7' 1 ~' ^,.f' .~7i?::i ~.?T3. i.~::.: i1 t ~~ .i 3 a ~,(s ? b y tit:
o TYI= r1"~ 1',.fll tr F~ t^ 9;~,, .i"r4J` )li _ .?il.: - c ,T ws most :,:f -k,.he Fav. i.,:i.an books
fi?~1ei7'~'r,
t"' ,;;, Fi P?Y1.L,2 ~ ,i'i.~.v~112. ~,r=; ~? .e:''~'} 7.Z,~. K )f
in tbiB ca a,1.oged b)r "he Ca." a?.1.Og'
''ra."'':n C'1f' t.J ) req. .i"4 F..""~' . . ? he :i " e.
Sep- - x. ...,. .. a._ _.
,~ _ tir, . ] r r C' 5 I) .t a r_~l1 t + ,7'tainaaYne l ; 7 0 h ::i y tY) S ,;
re
f ',~ J1. 1 F;;1 i7:C1
6lL'. ~! `~ N c ~ Tt 1 '1 ti? a i. d.lY r r' i,.Y.o.ige N .IS tin 4Th
i,'ti ^I C) r, ~,~ r"~'Q'.laC n G :?1? J -r 7 n:jf L _t-?, t111Sa1".'~.,.:" .,F;'~ 1 elSi and
r .T
e:rt',:r,a
of ) ,L" e FJ' A
l ~~rn ~~Ii.. 1,.4 F
a {'r -?r a.a ~'' I r- ")`a .....1 1~ j: 1~: + .? r!,.L t. t_ ?P. t..`." ,. ,1 .) l;tza'~S_ ." :t t,h"L `a a''~:J.Z.i.C~ N' ; 3 ,' P c1 r e
C. U. an
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6 !WA td
 J
Copy Approved for Release 2011/01/04: CIA-
SECRET
TTR/5 -12
the Reject
Bo Co.l.lection, and the: ore:ign Serial Collection (c .f 0 ,
Finding b i.nunediately be.lo~v) is d:uplicat,ed in the "M" Building
col.lectiors
b? Approx:imatt::ly 20 Russian serials are held
.
I~1"` Bull ding. vPM'" Building has found these particular. Russian
serials of such value as to require copies close at hand.
c. Foreign language bibliographic materials - in large part
duplicative ?-- are kept at both buildings. These materials are necessary
to the functions performed, in both buildings
do There is, of course,, dupl.:ication. of function in the case of
each of the above findings but no serious duplication of collect:i.ons
e. Duplicate catalogs on Oriental books received since 1953 are
kept at both "M" and
fG The Oriental Book Collection, Foreign Serial
Collection, and Reject Book Collection appear excessive in size for their
value and the use made of them. The Domestic Section Collection serves
its intended purpose but it, also, can. be reduced in size
ao When the CIA Library becomes centrally housed, the Oriental
Book Collection and. the Foreign Serial Collection,, as then constituted,
should be incorporated into the main Library holdings. This merger
will, of course, eliminate the question of overlap in these collection.:.,
and will obviate the need for the du.plicate card catalog on. Oriental
Books now kept
b? The recommendations which follow are essentially gratuit.pus In
that they do not fall. within the scope of the present, topi.c o They are
recorder' here, however, as of possible interest to OCR manag ment
1) The Oriental Book Collection should be reduced in size o _^..
very s:,bstantial redaction of tiLws collection be accompli he
without ser.ni.ous impairment of.' services th;rO"a; i disposing of
c. 18,000 7apanese be-c,ks receive-.d from the Washington, Docurment
These Japanese books are very rarely use-
l-(2) A reduction in size of the Foreign Serial Collection may
eventually be forced by considerations of space, iaanpower, and pro-
ctiv t
cur. ernen t, costs e The Task. Team accordingly feels some sele1)
on Russian and Or .ental serials (all such re::eipts are now maintained
in this col.lectic;n ) might prove advantageous. It is r econ. ended.
that uSe.rs of this collection be asked to suggest to the Library
areas of c~u: t?ailm rt;o ~:or..=ii.?:rat,io.n should also be given ~~o gr c-ate.r'
dependence upon the serial. holdings of the Library of Congress
SECRET
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 SECRET
TTR/5-13
Present CIA holdings Russian and, European Satellite serials are
largely duplicated at the Library of Congress. Retirement after a
fixed period of time o CIA serials now maintained in depth might
also prove feasible,
(3) It is recommended that competent personnel be sent from
consumer Offices to screen materials in the Reject Book Collection
in accordance with their respective needs. Materials selected by
this screen should. be given the requester on an expendable basis.
The remainder of the Reject Collection should be sent to the
Library of Congress for disposition. This screening by other
Offices should. be repeated periodically as the collection rebuilds
in volume. Disposal of new receipts via the Library of Congress
should be delayed long enough to cover most requests from consumer
use of the Russian Book List.
(4) The Domestic Section Collection should be reduced in size
by transferring all items carrying call muwbers either to the main
Library collection when space becomes available or to the Library
of Congress, as appropriate. (The Domestic Section Collection
would then be composed of "sterile", hard-to-get, and high interest
extra purchase items.)
5. Circulating Books and Periodicals
The Book and Periodical Unit, Circulation Branch, is responsible
for circulating Library books and. pericdicals and maintaining charge records
on items loaned.
The book and periodical collections held by the Acquisitions Branch,
however, do involve the Acquisitions Branch in circulation and charging
activities as follows:
If the book requested has not been cataloged,)
forwards the book to the Catalog Sect! on.9 Acquisitions Branch for
cataloging. (Exception.: uncataloged Oriental books are sent directly
SECRET
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
SECRET
to the patron.) A notation on a 3x5 card, however showing where the 25X1
item was sent and the date is made F_ -Il After cataloging,
the book is sent to the Circulation Branch where it is charged out to
the patron through normal Circulation Branch procedures.
Requests for books serviced by the small collection held in the
Domestic Section, Acquisitions Branch, are normally requests for retention
and the materials are sent directly to the patron on an expendable basis.
In those cases, however,, when Domestic Section materials are circulated
on an accountable basis, the items are sent to the Circulation Branch
for charging and delivery to the patron.
FINDINGS
It is true that the Acquisitions Branch is engaged in the circulation
function; duplication of effort, however, is not an important issue here.
The only facet of this functional overlap which might be construed as
duplication of effort is the 3x5 :card file :circulation.:notations::made by
on books circulated from the: collections . 25X1
RECOMMENDATION
When the Library elements have been collocated in one building, the
circulation and charging of all books and periodicals should be handled
by the Circulation Branch. (The 3x5 card file now maintained
will be unnecessary when this move takes place.)
6. Procuring Intelligence Publications
At present the following OCR components are engaged, as indicated,
in the function of procuring intelligence publications:
a. Circulation Branch, CIA Library, is responsible for the procure-
ment of documents on an ad hoc basis from the State Department
the Air Force. It is also responsible for procuring certain
documents
25X1
25X1
b. Acquisitions Branch, CIA Library, is responsible for the pro-
curement of Defense Department manuals, handbooks, technical publica-
tions, etc., available through the Defense Department and/or the
Government Printing office.
c. Document Division is responsible for effecting changes in
requirements for documentary series and servicing requests for CIA 25X1
material
d. Liaison Division is responsible for procurement of publications
from the Army, Navy, and non-IAC agencies.
SECRET
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6Ila
;~
 SECRET
TTR/5-15
It is evident that the functional overlap between these Branches
of the Library (the compass of this Task Team's responsibility) must
eventually be resolved within an Office-wide context rather than a mere
Library context. Considerable study has already been and continues to
be - given this problem by the top management of OCR.
FINDINGS
In that both the Acquisitions and-the Circulation Branches are
engaged in procuring publications from the Department of Defense an
overlap in function clearly exists. However, there is little, if any,
duplication of effort here in that the types of Defense Department
issuances to be procured by each are clearly delineated.
RECOMMENDATIONS
The Task Team, inescapably, thought about this problem to some
degree outside the limits of its responsibility - i.e., within an OCR-
wide context. The following recommendations result from a no more
than general examination of this problem but are given here for what-
ever value they may have:
a. The number of components engaged in the procurement function
should be kept to the minimum.
b. The Document Division should handle both the initial procure-
ment of government serials and changes in the numbers of copies
ordered.
c. All procurement involving the expenditure of funds should
be carried out by the Acquisitions Branch, CIA Library.
d. All procurement not,involving the expenditure of funds
should be centered in one OCR component... except as noted in b
above.
e. As a transitional measure, the following is recommended:
(1) that the Inter-Agency Unit, Circulation Branch, be
transferred (personnel and function) to,the Liaison Division
to serve as the nucleus of a central procurement unit;
(2) that procurement functions now carried out by the
Liaison officers be transferred, as feasible, to this central
unit;
(3) that this unit, at the time all non-expenditure OCR
procurement has been consolidated in this unit, be finally
assigned as most appropriate within the OCR structure as it
exists at that time - preferably to that component charged
with the over-all acquisitions function.
SECRET
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6~;,;
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
SECRET
TOPIC /r2
The work output for ordering and receiving
activities in the Domes tic Foreigri,~ and
Dissemination Sections, Acquisitions Branch,
is below standard.
SOURCES OF INFORMATION
i)
TTR/5-i6
Consultants' sources:
1.
A letter dated 10 May 1957
Asst Chief,
Library Division, Department of State
Appendix A).
3.
A paper titled "Work Load Categories
of Publications Procurement
for CIA", dated 7 May 1957
(Appendix C).
25X1
25X1
25X1
25X1
Additional sources: (Utilized by the Task Team)
1. Acquisitions Branch Monthly Reports for Fiscal Year 1956.
2. Data on foreign and special procurement from 25X1
Chief, Acquisitions Branch, CIA Library (Appendix D.
3. Data on domestic procurement from
Domestic Section, CIA Library.
Chief ,
4. "Title" and "item" classification for the tally listed in
Appendix A from Department of State, 25X1
Library Division.
DEFINITION OF TERMS
"Item received" - one physical unit of material received; i.e., one copy
of a book) periodical, newspaper, etc., received. (Receiving workload is
measured in items received.)
"Title ordered" ?- one bibliographic unit ordered; i.e., one complete
bibliographic identification recorded and dispatched for purchase action.
(Ordering workload is measured in titles ordered.)
"Subscription placed" .- an agreement to purchase one copy of each issue
of one periodical for a specified period of time. (Subscriptions placed do
not measure either orderi or receiving workload. They are more properly
commitmen~
a measure of monetary
SECRET
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
SECRET
INTRODUCTION
The Consultants stated that: "The output of work per staff member in
the Acquisiti.ons Branch is lower by at least one-third than that normally
found in research libraries., including comparable acquisition jobs such as
State Department Library,_" (Consultants' report, page 44, Finding #2).
(Although the State Department Library was stressed, the Consultants also
used the 'Yale University Library for comparative purposes.)
The Task Team studied the Consultants' method of computing work output
and the workload figures used.. We accepted the Consultants' basic method-
ology, in so far as it (1) undertook to measure work output in terms of
volume of material processed and (2) sought to compare the procurement
activities of other libraries, procuring almost exclusively for themselves,
with that portion of the CIA procurement effort conducted on behalf of CIA
recipients only. We did find, however, considerable distortion in. both the
Consultants' workload figures and the manner in which these figures were
used
Our comparative analysis of workload figures has been limited to the
CIA and State Department libraries for the following reasons:
1. the Consultants stressed. the comparison with State
3.
as computed by the Consultants;
the work output per staff employee was higher for State than
the procurement functions of State are more akin to those of
CIA than is the case
In compiling CIA Library workload figures, we have, where possible, u5,,4
the Library's official monthly reports as our data source. For items n'.t
covered in these reports, we have procured figures from the Chief, Acquizi.
tions Branch and/or the Chief, Domestic Section, Acquisitions Branch, as
appropriates Workload figures for the Department of State Librar were
supplied/
In so doing, these officers provided the Task
Team with a careful analysis and explanation of the workload, fig-7res
supplied to the Consultants n)the procurement activities of the State
Department Library (Appendix
This report (1) examines the Consultants' method and workload figures,
(2) recomputes work output per staff member for the CIA and State Department
libraries, using the Consultants' method with corrected workload figures,
and (3) computes work output per staff member for CIA and State using
corrected workload figures and the Task Team's method which distinguishes
between order activities and receipt activities.
SECRET
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
25X1
SECRET
P TTR/5 -18
THE CONSULTANTS ? METHOD A1VD WORKLOAD FICTt7R.ES
The method employed by the Consultants in arriving at their findings
was, very generally, to compare the procurement components of CIA's Library
with those of the State Department libraries and to
do so in terms of work units accomp is e per person per year. More speci-
fically, the Consultants' method was:
1. to define roughly similar procurement activities at each
library;
2. to compute work units accomplished by these activities;
3. to estimate manpower consumed in performing these units;
and finally,
4. to compute an annual work-unit figure per employee for each
library and to use these resulting figures as the relative measure
of work output.
The Consultants recognized that the acquisitions elements at
State are servicing the needs of their respective agencies only, whereas
CIA's Acquisitions Branch services a large number of government agencies in
addition to CIA. The Consultants sought to compensate for this difference
by limiting the CIA work processed and manpower input figures to that
portion of the Acquisitions Branch effort concerned with procurement for
CIA recipients.
The Consultants procured statements from each library of their FY 1956
work processed and manpower figures.' (The sources of these figures are
attached to this report as Appendices A. B, and C.)
The Consultants then divided, for each library, the total number of
items processed by the number of people involved. By this method, they
arrived at the following figures:
4764 units/per person/per year (15 persons)
4189 units/per person/per year (18 persons)
2614 units/per person/per year (26 persons)
The many differences of mission, operating procedures, and types and
volumes of materials handled in the libraries compared make it impracticable
to attempt numerical comparison of work output on any basis other than that
chosen by the. Consultants - namely, the volume of material processed. It is
felt, however, that the Consultants have arrived at an erroneous conclusion
because of both edata oomissions and errors in interpretation of the basic
data used in their
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6 `~yYq
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
J
J
SECRET
T
The error in .nt, rpr
reta,ti,:)n by the., Consultants bec:.aui; e the data
for State and c:: A not ex~x
f art.; .., in' ,.r? eased in like 4:L?7.'?.t~Fi . B"'~;.kt 7;7e St'~;~.Ge and. CIA
figures (Appe:r_, Aand, C} ;;rapp~.: ed t,~ the c iu c, *.,~;r w r'k.loa.d
figures e ,rc!.?gee.d. in units c;:t "items"v (r'ec:eipt units,, y a":f"titles- (order
Units), and of "s,cbscr I,ption_s" (mcnetar.7 'units, most pr, per'ly) ? These very
different types of unite, cannot, be intermixed and a..d:Sr. to one another on
a straight ari u'}1r~e a ~~' baw,1?s y as did the, Consultants 9 without invalidating
results ? For e:x:amp:ie, a. library mi.gh t order Life ra?gazinF at ti-,re different
times during the c.:UT;e: of a fiscal year with an. average of 3 copies
requested. in each order , In terms of ,units t: .his wouId repr'eeen t 5 ,titles
ordered, 15 subsc:r;ipt,ic,bs pla,ce.1. and 780 itemm received An intermixing
of such d.if`fer frig or,,~.t,s in aritbnietic computations clearly invalidlates the
resulting totals
To arrive at total work u:,n is y the Consultants sixaoiv aided t..:he figures
on State and CIA (Apgen,l.i c es, A~a?nd Cs r'espectiv el.y) Tn, i.tryc,ut regard to the
basic differences in the types of units given? ` hc-. fol.lc~'r.:ng indicates
something of the distortion caused by this action,
Most of the State Department work units fAppend.ix A) are expressed.
in 'litem r?eeceived." units. The figure on x i.r.,~3_i.c:2?: e %l2, 275 ~, hcwe~rer,
is a "subscription" tigur'e, which meas-are:s nei'?;ner? ordering nor
receiving workl: an.? Task Team consultations with State Department
personnel reveal that the ordering workload here was .,515 units
(includes newspapers) and the receiving workload figure was ::216,138
In that the: Consultants did not distinguish bet:err=:_r types of units,
we do not know which of these two workload ffig1ar'es tahe;;: might: have
pref erred 0 The .fact remains, however., that the. figure they t' id use
(subs cri vicns does not measlare workload for the. -or
receL1ing functions
In addition to the distortion resulting from f'ai,lure to distinguish
between the various t,ax-e:s of fig',xres which both CIA and State furnished to
the Consultants, there f.s t:he further problem that the CIA fl,gares: were
incomplete in terms of the use made of them., Although newspapers,, for
example, are reflected in the State figures, they, were not-, included in the
CIA tabulati.c,n? Yet, CIA both orders and. receives a huge vc-1,:me of news-
papers.
EEC! ONMYFOTAT ON OF' WORK O"-.q-P_ (Consultants I Method)
The Consultants' method of computing work c,ya.tp'a'i: per staff member was.,
as indicated abuvey (?;i,vid? manpc'we-r into a s.i:aglr fligtxr'ey
t ; t?
p,j_rportedl.y r. epresent:.if.g t+acquisitions run,c t .c>rt~"? The nn:~,,-t leg a.ma?te
single figure to use with such a method would seem to be the "7.tF'l s
? I. F: ~ , t+fze 'rneas'ur~e of the total. take: or ?payics,d'' resulting
received" i.
from the acquis,.t:i.c.,Ds
SECRET
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6 --A- 11111111110ift
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82TOO271 ROOO1OO12OOO4-6
SECRET
TTR/5 -20
The "items received" figures for acquisitions components of the State
and CIA libraries for F?.f. 1956 are: (see Appendix E for breakdown-)
State 2579279 items received
922,1+8 items received
These figures, when divided by the manpower totals used by the Consul-
tants for both libraries (State: 15 persons; CIA: 26 persons) give work
output figures as follows:
State 17,152 units/per person/per year
CIA 38,159 units/per person/per year
On the basis of this single-figure computation, then, the work output per
person in the Acquisitions Branch is more than double that of State.
COMPUTATION OF WORK OUTPUT (Task Team Method)
The Task Team has reworked this work output problem, remaining to
the extent possible within the Consultants' basic frame of reference.
However, to avoid the pitfall of adding together dissimilar types
of work units, we have distinguished between the ordering function and
the receiving function in our computations - a distinction the Consultants
did not make. This distinction, of course, gives us two sets of workload
figures and comparative measurements for each of these two functions.
Figures used, recorded in detail in Appendix E. were compiled as indicated
on page TTR/5W 17 above.
After consultation with State Department personnel, we have reducel the
State manpower figure to 13 (a reduction of 2 from the Consultants' figure).
We have used the same manpower figure for CIA which the Consultants used,
refining it only to distinguish between ordering and receiving personnel.
Comparative work output figures, computed as abc ve, are:
Ordering activity
State 1,966 units/per person/per year
CIA 3926+ units/per person/per year
Receiving activity
State 36975+ units/per person/per year
CIA 68,.21+ units/per person/per year
figures indicate that, in the ordering function, CIA personnel are
These
processing 66% more work units/per staff member/per year than their State
counterparts ..and, in the receiving function, 86% more work units are
being processed per employee in CIA than in State.
EGRET
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82TOO271 ROOO1OO12OOO4-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
SECRET
FINDINGS
1. The Consultants' basis for their charge that work output per
employee is scab-standard in the ordering and receiving activities of the
Acquisitions Branch, CIA Library, is invalid.
2. The figures computed, by the Task Team may be interpreted to mean:
(a) that CIA Acquisitions Branch employees are actually more
efficient in the performance of their duties than their State
Department counterparts;
(b) that a bias, attributable to mass handling of material by
CIA, results in speedier processing of material;
(c) that, since these figures represent per person averages
based on numerical totals for material actually processed, it is
possible that CIA employees simply had more work to do, and that
under a relatively comparable workload, State employees could
perform equally well;
(d) that a combination, of undetermined proportions, of the
three foregoing possibilities explains the differences'in'the CIA
and State output per person figures.
SECRET
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
SECRET
TTR/5 -22
TOPIC -',#r3
The Domestic 'ection should be moved
Tie principal objective in the minds of the Consultants in recommending 25X1
t.h,i.t the Domestic Section be moved from the Stadium to
=.C q?.Ii ri ~g of space in the Stadium into which the Reference and Circulation
Branches of the Library might be moved. This move of the Reference and
Circulation Branches ims at part of the Consultants' plan to establish a
central OC,1 information or customer service point. If it should be decided
to establish. such a central contact point in the Stadium (this concept is
ncu being studied by Task Teara #11), it may, of course, become necessary
that the Domestic Section be relocated.
Aside from the possible dictates of OCR reorganizations, however, there 25X1
is little reason to move the Domestic Section The principal
gain in such a move would result from having all procurement components of
the Acquisitions Branch located together. There would probably be. certain
management advantages in such a consolidated housing.
However, the Domestic Section is now more favorably located in relation
to (1) the rest of the Library, (2) the downtown book market which its
rersonnel roust visit daily, (3) the Agency components receiving the bulk of
its services, and (h-) the Machine Division which participates in its account-
in; controls.
An alternative move to accomplish the h sical consolidation of the 25X1
.cquisitions Branch would be to move the component.. to the
IIeaadqu.arters area. A very serious objection to this plan, however, arises
from the fact that this would separate the operation from its
rri.m::.ry customer for foreign publications ... the Foreign Documents Division,00. 25X1
in ac'd.ition, the elements of the Acquisitions Branch are
ui;i.qualy dependent upon FDD to supplement the foreign language capabilities 25X1
of their Personnel.
FIITDITY GS
(1) The advantage,; of the Domestic Section's present location in the
h those of its collocation with
i
t
g
we
learly ou
II_.?d U ,lt~_r5 area. c
T 1111~~1 17~ Sections .
(2) Tile aavn.iluci.6. v+ ,..~_~---
oroximity to the Foreign Documents Division outweigh the gain to be de-
olidating the Acquisitions Branch in the Headquarters area.
rived from cons
Rr~.CIIDIaTIQP1
T= s recommezid-e=d- that the Domestic Section not be moved to
S CRET
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
TOPIC ,4
Define the dissemination and distribution
junctions of the Acquisitions Branch and
the Document Division and study the
feasibility of merging like functions.
"Dissemination" -? the intellectual process of determining who is to
receive a publication. (This involves the scanning of text and the corre-
lation of contents with the needs of consumers.)
"Distribution" - the mail-room process of getting a publication into
the delivery system. (This may involve locating distribution symbols in a
file but does not require scanning for contents or knowledge of consumer
needs.)
A. NATURE OF THE DISSEMINATION FUNCTION (Acquisitions Branch and
Document Division)
1. Acquisitions Branch
Approximately 80% of the foreign serials, and 100% of the
domestic serials are received as a result of specific orders placed
by someone in this or the other agencies which the Acquisitions
Branch services. No dissemination is involved in this take because
the recipients are known.
Twenty percent of the foreign serials received are gratis
or sample copies, however, for which there is no previously estab-
lished dissemination pattern. Therefore, the area coordinator (in
the Foreign Section) disseminates the initial issue of gratis and
sample copy materials. Dissemination is made to Office, Division,
and/or Branch levels. The Acquisitions Branch area coordinator's
knowledge of consumer needs derives more from his handling of
purchase requests than from any formalized compilation of dissemi-
nation requirements as spelled out specifically for dissemination
purposes by the various Agency components. The dissemination of
these foreign serials, of course, requires of persons performing
this function a collective familiarity with a very wide range of
foreign languages- (To a very limited extent, serials are dissemi-
nated by the Dissemination Section although most of the handling
of serials by this section falls into the category of "distribution"
as defined in this paper.)
Offices receiving serials disseminated as above are requested
to indicate to the Acquisitions Branch their interest in receiving
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
a
TTR/5 -21?
future copies of similar material. In the case of sample copies,
the receiving offices are asked to submit subscription orders if
a continuing subscription is desired. From these actions, a record
of distribution is determined for future issues.
All books received in the Domestic Section result from
orders placed; therefore no "dissemination" action is required.
However, only 2C of the books received in the Dissemination Section
The disposition of these books is determined as follows:-
The Library Selection Officer and Screening Officer scan this
take for their respective interests. It is estimated that 25% of
these non-ordered books are selected by this scan for incorporation
into the Library an collections. Most of the remaining publi-
cations, with the exception of Russian and Far East materials, are
disseminated by the appropriate area coordinator, or by the Book
Unit, Dissemination Section, to offices known to have interest in
the subject and/or area content. Dissemination is usually based
upon the item's title and/or table of contents. Familiarity with
numerous foreign languages is necessary in performing this function.
The Russian and Far Eastern publications for the most part are not,
at present, disseminated
2. Document Division
Dissemination in the Document Division is limited to documents.
Dissemination in this Division is performed by professional
dissemination officers and is more highly organized than is dissemination
in the Acquisitions Branch.
Each Document Division disseminator has a disseminator's guide
book containing the reading requirements served on the Division by the
various components of CIA. Dissemination in the Document Division does
not involve foreign languages.
Documents to be disseminated are sent to these officers from the
Screening Unit of the Processing Branch, Document Division. The documents
are received in batches - a batch being from two to eighteen documents
of common source (State, Air, etc.) and packaged in an envelope. With
the varied requirements of Agency offices in mind, as recorded in the
disseminator's guide book, the disseminator scans each document in each
batch received. He arrives at a dissemination pattern for each document
based on its content. He records this dissemination pattern on a
"ladder-types card attached to each document. When the dissemination
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
SECRET
TTR/5 -25
for each of the documents of a batch has been determined and recorded,
the documents are placed into the envelope and the batch is sent to
the Processing Branch for distribution. Dissemination is made to
Office, Division, and/or Branch echelons.
B. NATURE OF THE DISTRIBUTION FUNCTION (Acquisitions Branch and
Document Division)
1. Acquisitions Branch
Kardex files are maintained by the Acquisitions Branch as
a record of both receipt and distribution for ordered serials.
The records also serve for follow-up work, reference questions,
and in the Domestic Section, for fiscal purposes. These files are
arranged in the Domestic Section alphabetically by title, and in
the Dissemination Section alphabetically by title within country.
Serials received are logged into the Kardex files and then
mailed to the distribution points shown in those files. Distribu-
tion of foreign serials requires the ability to recognize for
identification purposes serial titles and dates in many foreign
languages. To a considerable extent, however, this can be performed
through eye-familiarity with type-patterns and serial formats.
Distribution is effected to Office, Division, and in some cases to
Branch levels. Distribution of gratis and sample materials received
in "Y" Building is made in accordance with the dissemination deter-
minations of the area coordinators.
b. Books
Books received as a result of Agency-originated orders are
matched up with the Library purchase request form on which the
requester's name, office, and. room number are indicated. The flimsy
or carbon copy of this r:gae$t form s attached to the book and serves
as routing slip.
Books received as a result ofun-CIP eues are matched up
with the outside agency memoranda requesting ~c _
forwarded in accordance therewith.
LJ/~ I
Distribution of non-ordered books is
effected in compliance with the dissemination pattern as determined
by the area coordinators.
2. Document Division
The Processing Branch of the Document Division receives the
documents from the dissC:manators for distribution action. The documents
SECRET
MEMMUMMi Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271' R000100120004-6
 D
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
SECRET
TTR/5 -26
are d1stri:outed according to the dissemination card attached to each
report. Distribution is normally not made to any organizational
echelon lower than division.. Familiarity with foreign languages is,
of course, not required here.
FINDINGS
The following are the pros and cons, as we see them, of merging dissemi-
nation and distribution functions of the Acquisitions Branch with the
counterpart functions of the Document Division:
Dissemination Merger W Pro
1. Library and Document Division dissemination would be under a
single line of supervision. This might facilitate operational uniformity
and reduce supervisory overhead.
2. Dissemination of books would become more regularized. Books
would flow through the same dissemination "thought routines" as
documents.
3. The office reorganization effected to accommodate merging
disseminators might result in the separation of indexers from dissemi-
nators. This would disrupt the concept of interchangeability between
indexers and disseminators now stressed in the Document Division.
4. Books and documents are not easily compatible from the point
of view of physical handling.
Distribution Merger - Pro
5. All mailing operations would be under one line of supervision,
which should facilitate uniformity in mail handling, and possibly
reduce overhead.
6. All receipt and distribution files .. for serials would be in one
place. This Central. Srials file would comprise the Domestic and
Dissemination Section files, and the Document Division Standard Distri-
bution File.
Distribution Merger - Con
7. Coordination between ordering and. receiving personnel would be
more difficult because of their separation into different organizational
components.
SECRET
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6~
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
SECRET
rrR/5 -27
1. In terms of the present organization of OCR, we do not see
sufficient advantage in merging these functions to justify such action.
2. Whether or not such merging would be desirable under a different
OCR organizational structure is properly left to Task Team #17. We do
feel, however, that this merger concept, as regards present Acquisitions
Branch and Document Division activities, at least, does not merit per se
a significant influence on the formulation of OCR organizational changes
which may come under future advisement.
SECRET
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
Next 5 Page(s) In Document Denied
Iq
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
C-O-P-Y
In reply refer to
LR
TTR/5--Appendix A
May 10, 1957
According to your request of May 9th, we are forwarding the following
information on the Library Division' s work rates. These figures reflect
the situation during fiscal year 1956.
Acquisitions-.
ordering, receiving,
recording of serials, etc. Materials ordered and received consisted of
the following-.
1. Books and Pamphlets:
a. Ordered ............................ 6,132
b. Gratis .............................27,519
2. Research documents ....................... 7,381
3. Periodical subscriptions:
a. Ordered ............................ 2,Li11
33,1,,5/
b. Gratis.. ........................... 8,950 (estimated) i.2.. ,02, 7S
4. Newspapers ............................... 911
5. Microfilm reels .......................... 109
6. Foreign Service , etc .............. ...... 18,052
SECRET
S F CR'T
71,468
z76
9 s
7/, 4ti ff
4,76
11011111111111111 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
~:riu se
receiving spent about 20 of their time (1.L. .~e e pion aJ1U
positions) on ordered materials
and about 80% of their time on handling blanket order and gratis materials
and on problems relating to service, such as routing, inquiries about serials,
etc. Also, about two-thirds of the orders are not for the Library and
are not received by the Library.
Cataloging:
The hourly rate of cataloging for all types of cataloging is
2.411 titles and 4.52 items. Since, however, this is a composite figure,
we are also indicating the appropriate rates for each of the three types
of cataloging, viz.
25X1
25X1
Titles Items
1. New or origin
al cataloging.
1.78 2.24
2. Recataloging.
1.05 3.15
3. Additions cat
aloging..
4.118 8.18
Overall
rate 2.44 4.52
Interlibrary Loan:
The interlibrary loan function is performed by one position, and
the total transaction in FY-56 was 15,593 items. Of these, we borrowed
3,200 items and we lent 12,393 items.
Reference:
The reference staff consisted of 6 postitons which handled a
total of 36,532 reference requests in FY-1956.
Sincerely yours,
Assistant Chief, Library Division
5/10/57
S CR T
S CR T
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
Iq
Next 1 Page(s) In Document Denied
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
C
0
P
Y
1) . Books and Pamphlets
a) Ordered
b) Gratis
CIA Receipts - FY 1956
2), Research Documents
3). Periodical Subscriptions
a) Ordered
b) Gratis
4). Newspaper Subscriptions
TTR/S--Appendix D
Foreign Special Procurement
Orders Receipts Orders Receipts
3,013 3,013
2,611 2,611
3,321
3,321
390
390
4,980
4,980
735
735
5). Microfilm Reels 283
6). Foreign Service
7). Defense Publications
8). Miscellaneous (press Summaries,
press releases, publishers'
catalogs, posters, etc.) 109, 290 118,903
9) . Exchange (G.P.O. items and
other U.S. Governmental publi-
cations procured for trans-
mittal to Foreign Service
posts for exchange purposes.) 8,149 8,149
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
Iq
Next 3 Page(s) In Document Denied
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S-E-C-R. -E-T
CENTRAL,, INTELT_,1 GENCE AGENCY
OFF ICE OF CENTRAL RZT2-T-ENCE
FISCAL POLICY
TAS___ K TEA REPORT N0.
MEMORANDUM FOR: Assistant Director, Central Reference
SUBJECT:
26 March 1958
Final Report on Fiscal Policy for Publications Procurement -
Task Team No. 6
1? Membership
The Task Team consisted of the following members:
Li
As
Organization and Methods Examiner, Management Staff
aison Officer, Liaison Division., OCR
sistant Chief, Administrative Staff, OCR
2. Method of Task Team Operation ,
The following procedures were followed. by the Task Team in reviewing the
fiscal policy for publications procurement and evaluating the Library
Consultants' findings and recommendations pertaining thereto:
a. Review by all members of the Task Team of the Consultants' Report, the
OCR rebuttals, and other basic supporting documents.
b. Assignment of various phases of the problem to individual Task Team
members for investigation and reporting.
3.
c. Collection of data - the interview method was used for the most part
although considerable time was spent in reviewing files and records,
in examining procedures, and in studying published literature on
library operations. A list of persons interviewed is attached as
Appendix C C.
d. Discussion of reports from Task Team members, establishment of Task
Team position, and preparation of draft report.
e. Submission of final report to AD/CR.
ale recommendations contained in the body report are summarized as follows:
a. The publications procurement budget for FY 59 should be maintained at
$500,000. (p. TTR/6.3,4, & 5 )
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S'-E --C -I-, -h; -T TTR/6-2
b^ An overflow stack area should be established to handle
the period prior to the move to t' expansion during
, ,
c ? The experience of the NSA Library in movin to new,
investigated, as an aid,. in f g quarters should be
Building. (P. x/6?.8) P .ink the CIA Library move to the New
d. The CIA Librarian
regulation on should undertake to write and coordinate an Agency
the procurement of domestic and foreign books
and newspapers for Agency use. (p.TTR/6-lo) s Periodicals,
e. The Central Reference Advisory Group should investigate the system of
certifying purchase's within Agency components to insure that adequate
controls exist. (p. TS-.V/6-,g & I0)
f. The policy on budgeting for publications
Central Reference Advisory Group should be procurement as opted
retained. (p.aTTR/6-9)y the
g. The budget for expendables and. the budget for the main collections should
not be separate; the existing procedu e of centralized bud et re
and execution should be retained. ( g pr psration
tip :rTR/6-1011 & 12)
h. The Office of the Comptroller should be requested to provide technical
guidance in the setting up and maintenance of the fiscal records in the
Acquisitions Branch. (p. TIR/6-12 & 13)
i. The General Counsel and the Office of the Comptroller should be consulted
concerning the legality of the cash procurement procedure. (p.TTR/6-14)
-)
Chairman, Task Team Six
Attachments:
Task Team Report
A endix A - Policy on Budgeting for Publications Procurement 25X1
Appendix C - List of Persons Interviewed by Task Team Six
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
ne New Btulding. (p. TTR/6-5
67 & 8)
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S'-E-C-R?-E-T
TTR/o-3
PROJECT SIX; FT.SCAL POLICY FOR PUBLICATIONS PROCUREIvMNT
PROBLEM
To study all aspects of
To stt publications expenditure including the book
budgej, cash procurement, and fiscal, controls; specifically to determine
if:
A. The budget for publications is too low.
B. The rate of purchase is in any way affected by staff and space
limitations.
C. The present budget philosphy should be revised.
D. The budget for expendables and the budget for the main collections
should be separate.
E. The fiscal controls and procedures in Acquisitions Branch need changing.
II. TASK TEAM FRAME OF REFERENCE
A. The Task Team has assumed that it is CIA. and OCR policy to provide
the library services which the components of CIA may require, and,
to the extent possible, which other agencies in the intelligence
community may require, at a cost which meets the requirements of
national security as well as the requirements for economy in Government.
B. In studying the size of the publications procurement budget, the Task
Team confined itself to determining the adequacy of that budget to
fulfill the existing and known publication procurement needs of the
Agency, including the Library. The Task Team did not try to determine
the adequacy of the Library to fulfill the intelligence research and
operational needs of the CIA and the intelligence community.
C. The Task Team considered and used "book budget" as an inclusive term--
covering books, newspapers, and periodicals. We believe that the
Consultants used the term in the same manner.
D. The Task Team has proceeded on the basis that books, newspapers, and
periodicals are materials which are vital to both intelligence
research and operations. (The Consultants held this same view.)
IIT. Is the CIA Budget for Publications Procurement Too Low?
CONCLUSIONS
1. 500,,000 is more than ample to fulfill the estimated CIA
requirements for publications procurement in FY 58.
2. The $500,000 FY59 publications procurement budget is sufficient
and necessary to fulfill the FY 59 requirements of the program, as
presently estimated.
S-E-C-R-E-T
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R0001 00120004-6
ll.
S E-C-R -E -T
c. The term "normal research libraries"
compounds the fallacy of their com ' as used by the Consultants
(See APPEND C Part 3,) Parative library approach.
2. The Consultants, reasons for recommending that should be about $500.,004 Out of which OCR s ouldhs end aook budget minimum
$200,000 to develop its collections, are insufficient to suppo po of
the recommendation. (See AP END]x B, parts 2 .,3., 3, and k)
Task Team rai.sal of the FY 58 Publications Procurement Budget
3. The current $5009000 FY 58 publications
developed and approved procurement budget was
primarily on the basis of the Consultants,
recommendation. This sum is ample to fulfill the FY 58
_
tions procurement re uirements of the A ency in the pi ionlofa 25X1
the Task Team.
4. The CIA. Library is not now sPendin a
of the $500,000 budget to develo it
b? Comparing the
binding annual CIA Libra
ry expenditures for books and
books and
with such expenditures by university libraries and
comparing libraries,s
generally, without establishen their
basic similarities i income] ?~; tro
TM/6-4
B. FINDINGS
Task Team Evaluation
of the Consultants ? Findings
l? The evidence
which was offered and the methodology
employed by the Consultants to indicate that too ttle
is spent on developing the Libra
parison h Baas
money
com with ~ collections is based upon a
university libraries. Such libraries are not
truly comparable with the CIA Library.
a. The data dais, about CIA.
to develo Library expenditures for publications
p its collections which were cited by the Consultants
requires clarification and correction
25X1
onsultants.
collections" were defined by the C s colUections, as those
Task Team Judgment on the FY 59 Publications Procurement Budget
r
The $500-1000 FY 59 publicati-
procurement budget was the basis of previously established and time-tested procedures
on
w~~ich have now been formalized in CRAG document No. 2 58edure
These procedures are based on the principle that the
procurement budget is the sum in money terms of the Pubblicalcattons
tions
S-E-C-R -E -T
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 IM0016, Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S -E -'C .,R -E -T
superce e by mare recent editior..s, and outdated periodicals
5 scarded
are duplicate copies of books no longer in demand, volumes
d d
.
collections has been implemented. Inca~memora.ndum fromrthe
CIA Librarian to all major components of the Agency in October
19,579 it was requested that each component appoint members of
its staff to act as consultants in the selection program. This
program has had a promising beginning and is serving as a means
for utilizing the knowledge of all members of the Agency staff
in adding books to the collection.
The weeding program, which has been carried out in the Main
Library as well as the Branch Libraries, i8 a direct result
of lack of shelf space. Most of the materials beir A4
TM/6-6
c, Expansion Of +. stock space in presently available areas can
be made only be some Y by reducing reading area or office t e s
of Stacksnsta~use' of a fety regulations ti
Looms weight restrictions. the addition
.
d? A weeding program bas been and is being carried out
Mein Library and. the Branch Libraries in
1 the
for new accessions. in order to make e room
3* The results of the survey indicate that available s
Library is approaching ca prat pace in the
the Consultants that, the ce of FIowever, the assertion by
collections and the absence of certpurchase fr
ain basic materialssiso the
rimaril based on space limitations does not a
valid. An ana,jys; s of expenditures f
?r additions ~ Ps 't o to be
collections showed that the rate of the Library
constant for the p~St five A Purchase has remained nearly
conditconditions. Thies years in spite of increasingly crowded
would seem to indicate that the rate of purchase
was determined by basic Library of its collections Similarly, tereCisanotevidenceZto showothatt
absence from the shelves of materials that the Consultants
considered to he of basic importance is primarily the result
of lack of shelf ,spa.c:e. The determination of what materials are
of basic importance is made on the basis of established selection
criteria, and an e_,a,mi.mation of the selection program indicates
that materials are procured without regard
to space when deemed of
importance to the central collections.
4. The Consultants' claim that staff limitations affect the rate of
purchase cannot be substantiated, and no evidence was offered as
to how they arrived at this conclusion4
is determined by the Selection Section of tThe rate he Reference Branch
consisting of two Selections Officers. Increasing the size of
this staff would not, per se, increase the rate of purchase
since this is a result of basic Library policy.
5. The Consultants" recommendation that the Agency staff should
actively particir
te in th b
dispose of a. ,000 volume collection of captured Japanese
S -E -C,'R-'E -T
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
9.
S-E-C-.R-E-,T
TTR/6-7
A._ TTMli7e .v'ic14ncr
handleals.y ands, if successful, will.
their e.rsion needs gain enough space to
building In the opinion of thel~Task Team, thetweedin the move to new
Program is a normal and necessary g
eliminates material from the co ection thatwhashoutliv ed
Procere not its usefulness, but which
the value of the present holProvides
dings. a good means for judging
7? Although the Task Team wa,G
Clthough t s unable to substantiate the
finding that the rate of purchase of needed
books is affected by staff and s
analysis of the space limitations, an
anninued present stack spaces clearly indicates that
growth of the Library collection will be limited
by space considerations. Assuming that adequate space has
been provided for an e.
bee immediate expanded collection in the new building,
th
during next
e the Problem is how to provide for continued growth two-three year the new site. Several alternatives are prior to the move to
worth considering:
a? Emend shelf space in presently available areas by
rearrangment of the stacks to make room for more
shelving, utilization of reading spaces for stack
areas, and. continuation of a vigorous weeding program.
b. Establishment of an overflow stack area in newly
acquired spaces,
8. The Task Team has discarded the first proposal 0
not appear likely that enough space can be e ed by it does
method to handle the normal expansion rate of the collection
years. The most critical space
problem exists in the Main Library, which is
almost utilized, and immediate efforts should be made tain
more space to house this collection. The
Records
Staff is presently surveying the Fourth Wing of M Buildinggnt
to determine if more shelf space can be obtained by a
rearrangement of the stacks. Preliminary results of this
survey do not look promising. Conversion of reading room
space to stack areas-is not desirable and expansion of the
weeding program would in the long run be self-defeating.
The Task Team has concluded that the establishment of a new
stack area to handle the overflow from the Main Library offers
the best solution to the present space problem. Since space
is not available for a greatly expanded collection in M
Building, the following alternatives are suggested-
volume captured Japanese collection is disposed of,
there will be space to house an overflow in the existing
stacks. There is also a. vaulted area, of approximately 300
square feet presu:atiy being used as the Acquisitions
Branch conference room which could be converted to stack
spaces.
S E--C??R-E-T
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S L-C..R-E-T
T/6-8
b. Utilize existing spaces in the Stadium. The
Catalog Section moved out of an area of approximately
1500 square feet. Although this space has been
reallocated to other units, room could be made
available for a stack area in the Stadium.
c. Request the Office of Logistics to negotiate with
PBS for additional space not presently occupied by
CIA.
10. The decision as to what alternative to choose is dependent upon
an estimate of the space required in the next two-three years.
This estimate, in turn, is dependent upon the policy determination
as to what the rate of purchase shall be during this period.
11. One additional finding of the Task Team should be noted. In
discussions with NSA Library personnel, many of the problems en-
countered by NSA in their move from widely dispersed quarters to
a central building appeared similar to those which will be faced
by CIA in its move to the new building. One result of their
move was that a large number of books on-loan to user offices
were returned, thus greatly expanding the on-the-shelf holdings
of the central Library facility. This and other experiences
seem worthy of further investiga-ion by members of the CIA
Library Staff in planning the move to the new site.
C. RECOMMENDATIONS
1. An overflow stack area should be established to handle expansion
during the period prior to the move to the new building.
2. The experience of the NSA Library in moving to new quarters
should be investigated as an aid in the planning of the CIA
Library move to the new building.
V. Should the present budget philosophy be revised?
1. The OCR budget philosophy for publications procurement, as set
forth in the Central Reference Advisory Group issuance entitled
"Policy on Budgeting for Publications Procurement" (CRAG 2-58
dated 13 January 1958), is a clear statement of OCR responsi-
bility and procedures in this field. (See APPENDIX A)
2. The budget philosophy as adopted by CRAG establishes a satisfactory
method for the development and execution of the book budget.
3. The book budget is and has been under the control of the Assistant
Director, OCR. This is contrary to a finding of the Library
Consultants.
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S-E-C-R-E-T TAR/6-9
3.
B. FINDINGS
1?
In connnection. with
Inceonecionns with mplementing certain of the Consultants'
examination of the budget philosophy
for Publications pr urement was undertaken by Management
Staff and Acquisitions Branch officials prior to the formation
of the Task Teams. As a result of their investigations, a.
draft policy statement was prepared and submitted to the AD/CR.
After the Task Teams were organized, Task Team Six was requested
to review the draft and make any necessary changes. Minor
changes were suggested by the Team and the redraft was coor-
dinated with the O/DD/I, Office of the Comptroller, and
certain cognizant officials in user offices. The final
version was submitted to the AD/CR and adopted by CRAG at
its first meeting on 18 January 1958.
2. In connection with the OCR budget philosophy, the Library
Consultants stated that "the book budget is not under the
control of the Assistant Director, OCR, which is contrary to
normal research library practices." Although no evidence
was offered to support this statement, it appears likely that
the Consultants were referring to the procedure whereby OCR
requests user Offices to participate in the development of
the book budget by submitting estimates of their future
requirements. Since OCR has the responsibility for procuring
required books and periodicals as a centralized procurement
service and must bear the cost of these requirements, the
Consultants concluded that OCR does not have control of the
book budget. The statement of policy as adopted by CRAG
clarifies this point by stating that "OCR will be responsible
for the preparation of the budget estimates for the CIA
publications procurement program", and "will establish and
control the sub-allotments for the operating components".
The total funds for the program are allocated by the Comptroller
to OCR and the administering of these funds is the responsibility
of OCR officials. Should insufficient funds be available for
the operation of the total program, a procedure exists whereby
OCR may obtain additional funds from the user offices to meet
their requirements. This procedure is necessary only when the
total requirements for the program have been underestimated or
when the budgeted allocation has been limited by higher authority.
In OCR's annual presentation of the book budget, the Task Team
found that OCR has experienced difficulty in justifying the
program to the Office of the Comptroller. Most of the review
concerns the system of the control ("How many copies of the N.Y.
Times are bought and why?") rather than the relative importance
of books, newspapers, and periodicals to the intelligence effort.
The attitude of the Office of the Comptroller seems to be that
since user offices are not expending their own funds for
publications, there is a lack of restraint in requesting publica-
tions for office use; i.e., there is excessive wastefulness in
the program. The Task Team found that a control system does exist
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S ?-E - C -R WE-.T
TTR/6-lo
3-
in which an authorizing official in each Agency component
(Publicati.ons procurement Certifying Officer) reviews requests
for book pu.rchases and certifies the system
questionable feature, then their hsyste certifyinontrog l is
within Agency components should be reviewed to insure thatrthe
validity of the requests are certified by competent authority,
rather than attempting to control the program, by budgetary
limitations. The Consultants condemned this
process
edness and pointed out that when the total spent forabooks isight-
compared with the grand total needed to operate the Agency, the
sum becomes insignificant indeed." The Task Force concurs in
this finding,
4< Much of the difficulty encountered in the operation of the
Publications procurement program stems from a lack of understanding
by operating components of the purpose, scope and procedures of
the program. The CRAG issuance was an effective starting point
in clearing up misunderstandings and establishing policy and
procedures for the preparation of the book budget. As a follow-up
to this action, the Task Team feels that the publication of an
Agency regulation on the total publications procurement program
would be most helpful in giving the activity a firmer base from
which to operate, and in developing a clearer understanding by
Agency employees of this activity.
C. RECOMMENDATIONS
1. The CIA Librarian undertake to write and coordinate an Agency
regulation on the procurement of domestic and foreign books,
periodicals, and newspapers for Agency use explaining the purpose
and scope of the publications procurement program, detailing
responsibilities of OCR and user offices, and providing procedures
for publications procurement.
2. The Central Reference Advisory Group investigate the system of
certifying purchases within Agency components to insure that
adequate controls exist.
The policy on budgeting for publications procurement as adopted
by the Central Reference Advisory Group be retained.
Should the budget for expendables and the budget for the main collections
be separate?
A. CONCLUSIONS
Creation of separate budgets for expendables (in the operating
offices) and for the main OCR collections will not automatically
increase the funds available for the OCR collections.
0100011106-Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 "WAMEMIrs-anitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S -E-C -R-E-T?-
TTR/6-11.
2. Separating the budget r ,.or expendables and the budget for the in
collections Would further complicate the bookkeeping
the Acquisitions Branch, and would hamper the effectiveness ofsaof
centralized procurement service.
3. The procedure for developing the book budget as adopted by CRAG is
and., at the same time, maintaining ead quategcontrolsnt lization
B. FINDINGS
1. The Consultants presupposed a large sum of money, labeled "CIA
Publications Procurement" from which purchases are made for the
operating components, after which the remaining money is used
for the OCR collections. The Consultants believedthat if the
operating offices control their own funds for purchase of publica-
tions and reduce the volume of such purchases, the smaller amount
of expenditure, when subtracted from the total amount for "CIA
Publications Procurement" will leave a larger residue for use by
OCR to supplement its main collections. In fact, if there were
two separate budgets (i.e., one for each of the operating offices
and one for OCR), each would have to be justified in accordance
with normal budget jusitification procedures. Thus, OCR would have
to justify funds for its main collections in the same manner as
at present. Whether or not this would result in an increase in
such funds would depend solely upon the quality of OCR's justifica-
tion rather than upon the amounts requested by operating offices
in their budgeting.
2. The procedure whereby offices would prepare separate budgets for
their publications requirements could be effected in several ways:
a. The offices could set up separate allotment accounts for
publication procurement and authorize the Deputy Chief,
Acquisitions Branch to obligate against the authorization.
This would result in the establishment of 50-60 new allotment
accounts with consequent increased bookkeeping responsibilities.
b. The authorized funds obtained by offices as a result of their
budget justifications could be transferred to a centralized
allotment account controlled by OCR. If an office should require
more funds for publications procurement, it would be required
to transfer additional funds to OCR. Under the present system,
increased requirements in one office can be met from surplus funds
in other offices, i.e., the present system is more flexible.
c. The accounting responsibilities of the Acqu itions Branch could
be decentralized to the Office of the Comptroller and the Acquisi-
tions Branch act as a procurement agent only. This would result
in increased paper work and a slow down in the procurement
procedure that would greatly hamper the effectiveness of the
program.
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S-E-C-R E-T Ti'R/6-12
3. The statement of
procurement) policy and procedure on budgeting for publications
as adopted by the Central Reference Advisory Group in
J~Muary 1958 (see APPENDIX A), is a better system for achieving the
desired objectives than the system of separate
rs for fleexpendables
and the OCR collections. It has the advantages bofgg t
of accounting, and better utilization rofl brary ,
expertise resulting in a more economic administering of the program.
If the procedure seems too cumbersome, it is because of the requirements
of higher authority for adequate controls and sufficient justification of
the program.
C C. REC MENDATIONS
1? The budget for expendables and the budget for the main collections
should not be separate. The existing procedure of centralized
budget preparation and execution should be retained.
VII. Do the fiscal controls and procedures in Acquisitions Branch need changing?
A. CONCLUSIONS
1. The Management Staff survey of the record-keeping procedures for
publications procurement, as carried on by the Deputy Chief of
the Acquisitions Branch, has succeeded in partially reducing the
burden of record-keeping in the Branch.
2. The fiscal controls and procedures should be further studied by
Agency experts in the field of accounting.
3. The cash procurement procedure is being utilized to its fullest
practical extent under present operating conditions.
1t. The legal status of the cash procurement operation needs
clarification.
1. The Consultants recommended that "the proposal- of the Management
Staff to undertake a detailed study of fiscal control and
bookkeeping as now practiced by the Acquisitions Branch" be
implemented. This was accomplished in part in the summer and
fall of 1957 and culminated in the publication of a Management
Staff Report entitled "Accounting for OCR Publications
Procurement` dated 14 October 1957. The objectives of the
study were:
a. "To reduce the burden of the Deputy Chief, Acqusitions Branch,
CIA Library, in managing ~he fiscal records for Agency
publications procurement.
"To devise a more efficient subscription renewal procedure
for foreign publications and domestic annual publications."
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S...E_C-?B: B T
MR/6-13
2. As a result of this study, members of the Management Staff have
been working with members of the Acquisitions Branch to implement
certain of its recommendations. The main accomplishments to date
are as fnl l .,T.,~ ..
has been delegated from th
e Deputy Chief, Acquisitions
Branch to the Chief, Special Procurement Unit.
b. A procedure has been devised and. is in the process of being
installed for machine accounting for foreign subscriptions.
c. Steps are being taken to improve the procedures for machine
accounting for other foreign procurement
d. A more economic system has been devised for creating and
maintaining the cyrillic list of Russian. publications.
e. Some accounting reports previously prepared manually by the
Deputy Chief,, Acquisitions Branch, and Chief, Domestic Section
have been converted to automatic machine methods.
3. Although improvements have been effected in. I&M accounting systems
for the Branch, the major problem of simplifying manually kept
fiscal records has not been completely solved, and the Deputy
Chief, Acquisitions Branch, continues to devote a major portion
of his time in preparing and maintaining budget and fiscal records.
Guidance from the Office of the Comptroller is required to perform
a technical analysis of accounting operations in this Branch. This
would have the dual. advantage of improving the bookkeeping procedures
and of establishing closer contact between the Office of the
Comptroller and. the AcTx.sitions Branch.
4. In their study of the acquisition program, the Consultants found
that "full utilization of the cash purchase procedvln-e is not made"..
and recommended that "'more books and, other ma.teriai, .,e urchased
throw the cash Procurement procedure".]
25X1
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
.S R
6. Cash procureme_r
has One Min n cs.il:c .rztA c le s
of domestic g ~ y'`~`i t;oe. The a.:Ltc rms.tive method
:. .~.
tfine mes PYip(.utment by the pur:th:.se. ord.r procedure, takes more
larger typing people e caia e of,
tae great volume of paper involved,
.d,9 and. more detailed accounting
The main. eleariera't:s of ' the cash ~ ~ procedure.
h pr.?ocurtment system are
a. Cash is Mde ~,;c>ra,i, L b1.e to the Domestic Section through the
use of an iztvouc:hered. revolving fund s;moixn'ring to $1 O0.
b. Its use is limni.ted, to items a;vs.i7.s;h;t.e in the local book
market
c. Requi.remerts, -~." law e,?;:
of this, ;'rnt-'e.r%:.7.'F. t. Agency l.i?7it greater use
7? In their study of the cash
procurement opex::a.ti,~n, the :t .rsagement Staff
questioned the procedure on the grounds that Agency regulations
prohibit the use of unvouch.ered funds fo:r "administrative convenience".
Since they feel that the use of this fund. is m'An"iy a matter of
"administrrrti?u?e cornvvenjence." 9 they a, ^e pr?p::,sing; that it be changed to
vouchered funds by either.
a. Abolishing the revolving fund. and, estahlishirl.g a modified version
of the purchase order system, or
b. Establishing a voucheredd. ymprest fun.: as a replacement to the
unvouchered revolving fund..
8. The Task Team feels that the abolishing of the revolving fund and
establishing a purchase order procedure is not a good solution
since it would increase the work load. of the Domestic Section.
Establishment of a vouchered imprest f and ?w?oul d retain all the.
advantages of the present system and. resolve the legal questions
involved. However, Agency ? regulations limit imp'~rest funds ds to
$500 although "exception to this, limitation may be requested with
justification or the basis of the particular situation involved!'.
Since no less than. $4000 is needed to r9.in the c:,;peratiorl, an exception
to the regulation would have to be obtained from the Office of the
Comptroller in order to set up the ,fund.. No efforts have been made
by OCR or members of the; Management Staff to discover if such an
exception is possible. Inasmuch as the legality of the procedure
is in doubt, some solution to the problem shou-ld he made in the
near future.
C. EECCENDATIONS
The Office of the Comptroller should be requested to provide techni-
cal guidance in the setting up and. m:,iat e.,:,a.:,.c.e of the f'iseas. records
in the crl B man h ~,wi t ?h the., 1r,.;~ r r o' i ~. L+
A_ u.i.f~iti u!i' rv i -to
e.,iei; e the
Deputy Chief, .Acciu.i. .itions B.-r*,..ch .: f soyrn of his bookkeeping duties) .
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82TOO271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S-E-C -R -D, -T TTR/6-15
The General Counsel and the Office of the Comptroller should be
consulted concerning the legality of the cash procurement procedure.
it be determined that this is a misuse of unvouchered funds,
steps should be taken to establish a vouchered Imprest Fund of 1,000.
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82TO0271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S-E-C-R-E-T
TTR/6--APPENDIX A j
CRAG 2-58
13 January 1958
C E N T R A L I N T E L L I G E N C E A G E N C Y
CENTRAL REFERENCE ADVISORY GROUP
Policy on Bucr.geting for
Pub tail tions Procurement
At its 7 January organization meeting, the Central Reference
Advisory Group adopted the attached statement of policy on budgeting
for the procurement of publications.
I have directed responsible OCR officers to implement it, in
consultation with responsible officers in the various components of
the Agency, at the earliest practicable date.
Paul A. Borel
Assistant Director
Central Reference
DD/I
DDS
IADIS
Chief, FI
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
7 January 1958
POLICY ON BUDGETING FOR PUBIJCATIONS PROCUREMENT
1? General.
The Office of Central Reference is responsible for providing books,
periodicals, and other publications required for use by all components
of the Agency) Development of the publications procurement program,
including the preparation and execution of the budget, will be accomplished
by OCR after coordination with CIA operating officials.2
2. Budget Preparation:
a. OCR will be responsible for the preparation of the budget estimates
for the CIA publications procurement program for all phases of the budget
S -E -C -R-E -T
TTR/6--APPENDIX A-2
cycle (preliminary estimates, operating budget, revised estimates). These
estimates will be based on:
(1) past expenditures as reflected in the records maintained by OCR;
(2) known trends in the publications procurement program; and,
(3) new or discontinued requirements of the Agency.
b. Preparation of the budget will include separate estimates for
operating components of the Agency, the OCR central collection, and the
foreign publications selection program.3
2Operating officials include: Chiefs of Senior Staff and Area Divisions
under the jurisdiction of the DD/P; Assistant Directors under the juris-
diction of the DD/I; Staff Chiefs, the Comptroller, the General Counsel,
and Directors of offices under the jurisdiction of the DD/S; and the DD/C.
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6 ~'~
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
TTR/6--APPENDIX A-3
S-E-C-:R-E-T
c? OCR will be responsible for the coordination of individual estimates
for operating components with CIA operating officials at each phase of the
budget cycle (see attached sample correspondence). Operating officials will
propose revised estimates, if necessary, based upon anticipated changes
in their requirements.
d. Submission of the total Agency publications procurement budget
will be made by OCR through the Deputy Director (Intelligence) to the
Comptroller.
e. The Comptroller will allocate the approved Agency publications
procurement funds to OCR. OCR will establish and control the sub-allotments
for the operating components and notify them of the approved amounts.
3. Budget Management-
a. As publications are ordered by operating components, OCR will
charge the costs to their sub-allotments. Publications selected by PPO's,
or ordered for the OCR collection will be charged against the OCR sub-
allotment.
b. When the funds for an operating component are nearing exhaustion
by purchases against its sub-allotment, the component will be notified.
On exhaustion, additional. purchases for that component will be at the
option of OCR, unless the component makes supplementary funds available
to OCR.
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S E-C-R-E-T TTR/6--APPENDIX A-4
DRAFT form letter
(____February 1958)
MEMORANDUM FOR:
ATTENTION:
SUBJECT: Publications Procurement Budget: Sub-Allotment
for
The following report of sums involved in procuring publications
required by your component during the first half of the fiscal year is
submitted for your information.
Budget sub-allotment
Newspapers, 12 months
Foreign subscriptions ordered
Other subscriptions and books,
July-December
Subscription renewals due,
January-June
Total funds committed
Balance available for new selections,
January-June
Paul A. Borel
Assistant Director
Central Reference
S-E-C-R-E-T
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6 1>a
S-E-C-R-E-T TTR/6--AFPENDDC A-6
DRAFT form letter
MEMORANDUM FOR :
ATTENTION:
SUBJECT: Publications Procurement Budget: Sub-Allotment
for
1. Expenditures and net obligations to serve the requirements of your
component for publications during the fiscal year just completed total $ ?
2. The budget sub-allotment to serve the requirements of your component
for publications during the current fiscal year is $ ?
Paul A. Borel
Assistant Director
Central Reference
S -E -C -R E -T
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
25X1
Next 10 Page(s) In Document Denied
Iq
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
S -E -C -R-E -T
1 4
TTR/6--APPENDIX c
LIST OF PERSONS IlVTERVIEWED BY MEMBERS OF TASK TEAM SIX
Librarian
Deputy Librarian
Chief, Acquisitions Branch
Deputy Chief, Acquisitions Branch
Chief, Domestic Section, Acquisitions Branch
Chief, Cash Procurement Unit Domestic Section, Acquisitions Branch
Chief, Purchase Order Unit, Domestic Section, Acquisitions Branch
Chief, Selection Section, Reference Branch
Assistant Chief, Selection Section, Reference Branch
Chief, Catalog Section, Acquisitions Branch
Chief, Services Section, Circulation Branch
Executive Officer
Office of the Comptroller
Acting Chief, Budget Division
DD /P
Area Case Officers
Department of State
Librarian
Deputy Librarian
National Security Agency
Deputy Librarian
Chief, Collection Division
Chief, Acquisitions Branch
Department of Agriculture
Librarian
Chairmen and Members of Task Teams Four, Five, and Seven
S-E-'C-R-FE-T
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6 ~?~??~?INA
-+~+>
 Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6
Iq
Next 1 Page(s) In Document Denied
Sanitized Copy Approved for Release 2011/01/04: CIA-RDP82T00271 R000100120004-6