LETTER TO DR. WILLIAM T. KNOX FROM CHAIRMAN
Document Type:
Collection:
Document Number (FOIA) /ESDN (CREST):
CIA-RDP80B01139A000500050001-3
Release Decision:
RIPPUB
Original Classification:
S
Document Page Count:
65
Document Creation Date:
December 12, 2016
Document Release Date:
April 9, 2002
Sequence Number:
1
Case Number:
Publication Date:
July 8, 1966
Content Type:
LETTER
File:
Attachment | Size |
---|---|
CIA-RDP80B01139A000500050001-3.pdf | 4.94 MB |
Body:
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
July8
Dr. William T. Knox
Office of Science and Technology
Exeeutive Office of the President
Washington, D. C.
Dear Bill:
We take great pleasure in transmitting to you the
enclosed draft of the report of the subpanel on biographies
of the Guidance and Evaluation Panel.
This report cannot claim to be final. Indeed, as
we see it, it is only a bare beginning.
We wish to take this opportunity to thank those members
of your staff, and of the staffs of various community agencies
whose help was invaluable, not only in terms of excellent techni-
cal presentations but also in guiding our way through the lab-
yrinthine aftinistrative obstacle course the community presents
to the innocent outsider.
ely yours,
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
25X1A
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
DRAFT
T OFORN CONTROLLEDDISSEM
SECRET
Presideflt'a Foreign Intelligence Advisory Board
Special Assistant to the President for Science and Technology
Guidance and Evaluation Panel
William T. Knox, Chairman
DRAFT
REPORT OF THE SUBPANEL ON BIOGRAPHICS
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
25X1A
Approved For Release 2002/05/09 : CIA-RDP8SME CO
DRAFT
SEC CONTROLLED DISSEM
BIOGRAPHICS St3BPANEL REPORT
Table of Contents
Introduction
Summary and Conclusions
A. Background
B. Major Findings
C. Major Recommendations
II. The Biographies Problem
A. Information Sources for This Report
B. Subpanel Qualifications
C. Definition of Biographies
D. Significance of Biographies
E. Technical Simplicity of Biographies
F. Biographies as a Laboratory for Tec
III. Observations
A. Relative Characteristics of Positive Intelligence,
Counterintelligence and Security Files
B. File Overlap and the Single File Problem
C. Evidence of Current Automation Efforts
D. Need for Research and Development
E. The Nature of the Biographies R & D Problems
F. The Invisible Colleges
G. Compartmentation and Security
Ot014-3
Page i
and Management
Approved For Release 2002/05/09 : CIA-RDP80B01139A000500 500 t-34
Approved For Release 2002/05/09 : CIA-RI:130E1C
DRAFT
RET/NOFORN/CONT OLLED DISSEM
Table of Contents (continued)
107001-3
Page ii
H. Conceptual and Managerial Problems
I. Manpower
J. Communications
K. An Approach to Security Problems
L. File Conversion
IV. Bibliography
V% Appendices
1. The Nature of the Biographies Problem
2. Request for Information and CODIB Staff Response
3. National Driver Register Service
4. Classification of Biographic Intelligence Activities
5 Biographic Intelligence Systems Improvement
a. Evolution of a National Agency Check System
b. Search Strategies in Large Biographic Systems
c. File Conversion Problems in the Intelligence
Interagency Security Name Check Activity
_P !la
Approved For Release 2002/05/09 : CIA-RDP80B01139A000500050001-3
Approved For Release 2002/05/09 : CIA-RDOr IdA 0500051.001-3
DRAFT
3C RET OFORN CONTROLLED DISSEM Page 1
INTRODUCTION
The subpanel on biographies, like the whole Guidance
and Evaluation Panel, faced the question of how to provide
increased technical support for the information processing
activities of the intelligence community. The technical
problems presented themselves as inextricably woven into the
pattern of information use and therefore into the structure o
management and management policy.
The rapidly increasing growth rate of information in
the national data base obviously generates, a need for addi-
ott?.-.,X-
tional resources to aid in processing/this information. However*
since the potential information pool is practically infinite,
the rate of improvement in information processing is directly
dependent not only on the rate at which new technical approaches
are implemented, but also on the growth of management recogni.
tion of the fact that information processing policy must be one
of its basic concerns and that this policy must take into
account the need for selectivity created by the evident hope-
lessness of ever marshalling "all" relevant information.
The intelligence community is fortunate in having
extensive resources of modern science and technology available
for use now. The rate of advance in information handling
technology is itself increasing. Moreover, the community
already has a strong capability in applying automatic data-
processing techniques to cryptanalytic and communication
Approved For Release 2002/05/09 : CIA-RDP80B01139A0005090q00010
Approved For Release 2002/05/09 : CIA-RDPSOIr"3
C"RT71-3
DRAFT
0 RN/CO OLLED DISI
Page 2
problems. The community is therefore in a unique position to
apply these resources before the magnitude of the data base
outstrips human capability to process it effectively.
One area that seems amenable to the application of
modern data-processing technology and methodology is that of
biographies. This subpanel has studied the suitability of
automatic data processing for solving some of the problems
found in the biographies area. In the process, we have
examined biographic files both in and out of the intelligence
community. We conclude that given strong management policy
tAseystt
guidance, support and evaluation, graphics could benefit
greatly.
Approved For Release 2002/05/09 : CIA-RDP80B01139A000500050001-3
Approved For Release 2002/05/09 ? CIA-RDNOB0,1139AQQ05E0T001-3
DRAirr stuR
SECRET/NOPTIN/CONTROLL Page 3ED DISSEM
A.
ound
document is the report of the subpanel on
biographies of a Guidance and Evaluation Panel created under
the Special Assistant to the President for Science and Technology
and the President's Foreign Intelligence Advisory Board, in
response to a memorandum to the President by the Chairman of
the PFIAB (CLIFFORD)*.
The Clifford memorandum charged the panel with
responsibility for:
A. "Providing guidance to the intelligence community
in the forwarding of methods and facilities for
information handling and access;
B. Evaluating in technical terms the true meaning of
the enormous and somewhat heterogeneous growth of the
intelligence community's information pool."
The memorandum then goes on to say "It is emphasized
that the proposed panel of technical experts would not be
tasked with the too-obvious assignment of simply applying
modern machine methods to the existing* specialized, and
rigidly-maintained activities of processing and distributing
information within the intelligence community. The panel would
have the overall task of guiding the necessarily large, and
Capitalized expressions refer to the corresponding entry in
the bibliography.
Approved For Release 2002/05/09 : CIA-RDP80B01139A000500050001-c
Approved For Release 2002/05/09_._. CIA-RDP804?Q014.39A_Q00500050001-3
DRAFT NEGRETPage 4
/X0FORN/CONTROLLED DISSEM
presently ignored, planning for the realisticand long term
development of mechanized facilities for the processing of
information in the manifold forms in which it is encountered
within the intelligence community."
Adhering to this spirit, the subpanel has not
hesitated to seek information outside the confines of the
intelligence community narrowly interpreted as participants in
the United States Intelligence Board*
The panel was given an extensive series of briefings
and site visits in the major intelligence installations in
the Washington area and elsewhere and access to some documents
produced by earlier studies. (LISTING TO BE nicL)DED1
It quickly became clear to this subpanel that the
whole subject area, and biographies especially, has an excellent
history written not only by innocent outsiders but mainly by
very able, knowledgeable, conscientious and thorough members of
the community itself. Since the Kirkpatrick Report
(KIRKPATRICK, 1960), more time has been spent, more detailed
data has been gathered, more debate and interpretation has
taken place than any part-time "expert however well
Intentioned, could possibly equal.
Without apology, this report therefore makes very
little claim to originality. In the words of Tom Lehrer,
"Everything, I stole from somewhere else," including of course
information gathered from briefings and some more informal
on-site spot checks of critical items.
Approved For Release 2002/05/09 : CIA-RDP80B01139A00050005000-3
Approved For Release 2002/05/09 : CIA-RDF.1r
DRAFT
PrOFORN/CONTROLLEDDISSEM
I II II
a
5tplooi-3
Page 5
B. Major Findings
The major finding of this subpanel is that the main
thing preventing an immediate intelligent communitywide attack
on information problems in the biographies area is a combina-
tion of inability and unwilliAgneep or the intelligence
Go Nre ,400410%
community4to pay more than lip service to the idea of concerted
effort. It is the so-called "realities" of organizational
politics, not any basic technical obstacles,that spawn apparent
impossibilities at every step of the way. No serious technical
obstacle is visible on the way toward automation in the
biographies area. We do not doubt that technical problems will
appear further on, but the way is clear enough to set out on
the journey at once.
C. Major Recommendations
Our major recommendation is that a clear-cut locus of
management responsibility and authority for effective decision
and concerted action be created to:
1. dispel the fog that sweet talk of unity casts over
a fundamental parochialism that precludes intelligent
collective action.
2. focus attention on basic problems of research and
development, technical feasibility, systems analysis and design
and management policy and control instead of frittering it
away on arguments over such profundities as the matter of how
many columns of a punched card should be allocated to a name.
Approved For Release 2002/05/09 : CIA-RDP80B01139A000500050-0014
r
Approved For Release 2002/05/09 : CIA-RDPigt3t00a0t571-3
DRAFT
ECRET/NOFORN/CONTROLLED D Page 6
3. establi8h technically competentlines of management
to be supplied with rational system performance and cost data ?
now almost totally lacking4 n order to guide e selection,
?1? the
collection, storage and retrieval of information in an
environment where the ideas of sampling and statistical confi-
dence must overtly displace the vain ideal of completeness
already abandoned in apparently random practice.
l4. insure that the sound, valuable but fragm ntary ideas
generated -- but not always presented -- in earlier internal
community studies be given a genuine hearing leading to the
vigorous exercise of management decision-making responsibility
In place of the current committee practice of making recom-
mendations reduced to the lowest common denominator.
5. that the results of isolated experiments be broadcast
withinAKC community, evaluated critically and impartially and,
LA
when found valuable, applied in response to vigorous management
direction.
6. ensure that the needs of users are taken into account
SA-L. ??L2.6.ib)
by drawing users actively into the earliest ages o systems
analysis and design.
The remainder of this report aims to make more
to detail and amplify, end to document these basic
ci
nions.
-7, -cc, clitAit
\)-:o ?,.:5,4-1L.Q_
? ?),z- vv4kk k?
et,.$ a-
sawsLc L-
tev,429. tatt-*J-441- 5
4,
A'ppr ved
20VJA.109-. CIA-RDP80601139A000500050001-3
L-
Approved For Release 2002/05/09 : CIA-RDP80601 gAgoboopoot3
SECRET 0 CO 0 DIS
Page 7
THE BIOORAPIICS PROBLEM
A. Information Sources for This Report
Throughout this report we shall have occasion to refer
to the report of COMB Task Team 5 in a draft form dated
January 17, 1966 (COMB Although than draft report does not
have the blessing of COMB as of the tme of this writing, it
nevertheless reflects the concentrated effort of a group of
twelve members of all components of the intelligence community
in gathering and evaluating data concerning the scope of
biographic files, the problems associated with creating, main-
taining and using them, and, in the words of the statement of
objectives of the team "to identify means for improving the
storage, retrieval and exchange* of information from the major
name files and related data files in the intelligence community."
As of 31 December 1965, members of Task Team 5 "have reported a
total of 3,993 hours devoted to this effort." (COMB 1966)
The subpanel is most grateful for the opportunity to use the
fruits of such extensive labor as well as other related docu-
ments detailed in the bibliographyllo-41
There is a major gap in the work of the Corfu Task
Team 5 which, unfortunately, this subpanel has lacked the time
and resources to fill. Although one of the strengths claimed
for the COMB approach" is that "participation by officials
derlino.
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05/09 : CIA-RDP80B0g3rcart.
DRAFT
SEC AOFORN/CONTROLLED DISE= Page 8
with operating responsibilities gives greater assurance of
utility and aceeptance of results" (BORZL 1966, p. 6), the
report or Task Team 5 is restricted to technical analyses that
give no evidence of participation by any official with
responsibility for seeing to it that the needs of file users
are indeed met. The notion of "operating responsibility
thus seems to have been narrowly construed to mean responsibility
for operating the files. Since it is now generally accepted
that sound systems analysis requires paying attention to the
users (and users paying attention to systems analysis), we
see this limitation as a major shortcoming of the CODIB approach.
The impression that this limitation is not mere
accidental oversight is reinforced by the following quotation:
"The SCIPS study group and the panel of consultants
recommended the establishment of a permanent inter
agency body of sufficient size to cover the problem
areas. COD IB however has rejected this approach in
favor of a number of ad hoc panels to be supported
by a small permanent executive secretariat (group of
documentalists). CODIB not only found excessive*
the time necessary to build up a sizable competent
permanent group (unless this were done at the expense
of promising intra-departmental programs), but
considers counterproductive any Plan which calls for
solutions developed by specialists not themselves
involved in departmental information operations"
BRIGGS 1964).
Since the directive creating this panel requires it to
"supplement the longer range Task Force project;being pursued by
the USIB (CLIFFORD page 2, paragraph 5) the availability of a
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
OFORN CONTROLL
SECRET
Page 9
report of the relevant Task,For e even in draft form is a
major asset. It is exceedi ly rare for a panel of "outside
A
experts" to have access to such timely data on its subject of
concern, produced by those most intimately acquainted with the
problem. The findings of the subpanel are based in addition
on visits to the major files [ADD LIST HERB' and on discussions
with those concerned with their maintenance and their use, as
well as with the COMB
C:
well We have also specially
requested presentations of certain specific technical facts and
opinions (Appendix 2) and these are attached as Appendices 3,
4, and 5.
We are sensitive to the qualms of the community
about the intrusion of outsiders. One CODIB Task Team urges
the acceptance of its recommendations on the quaint ground that
otherwise, the intelligence community "w 11 continue to be
vulnerable to external Investigative and evaluative groups
without having any recognized negotiating position" (COBIB VI
p. 11), further warning that such groups, however well-
intentioned, will tend to gather data that are fragmentary,
without interrelation ani.Xl therefore tend to make recom
mendations that are even an a related to real
problems. The warning concludes that n the process, the
outsiders will remove a thorn and, by so doing implant a
tumor" (CODIB VI p. 11). We do not feel immune to such
=;- but, by relying extensively on the community's own docu-
ments, we hope to present it with a mirror in which it will see
itselley3rmoteditor Release 2002/05/09 : CIA-RDP80601139A000500050091-3
E
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
DRAFT
SECRET
SEcRQ!O/CO!!Z DISSEM Page 10
B. ubpanel Qualifications
Although the members of this subpanel are clearly
outsiders to the intelligence community, their qualifications
have nonetheless been acknowledged by the community.
The Fifth Annual CODIB Report (CODIB 1963) quotes
from a reporte an ad hoc study panel of PSAC on non-numerical
information processing comiiiiia?in mid lgg)
Se SCIPS report includes the
owing recommendation (CODIB 19640 pi 31 Recommendation 3):
technical review panel such as the PSAC ad hoc study
panel on non-numerical information processing be assembled to
review the detailed findings of the study and provide comments
thereon." This subpanel consists of two-thirds of the afore-
mentioned PSAC ad hoc study panel on non-numerical information
processing. One of us was in the technical review panel for
the SCIPS report and the submission of this report further
implements at least this one
The March, 1964, "Progress Report on Foreign
Intelligence Objective No. 3' in mentioning the
establishment by the National Academy of Sciences of an
Advisory Committee on Automatic Language Processing comments
that "This disinterested group is to advise the Department of
Defense, the Central Intelligence Agency and the National
Science Foundation on Research and Development in the general
field of mechanical handling of languages ..." One member of
recommendation.
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-
25X1A
Approved For Release 2002/046?MIA-RDP8SOE GO
SECRET/NOFORN/CONThOIiLED Dissgm
E0C11-3
Page 11
this subpanel was an active participant in the National
Academy Advisory Committee.
Since, as will be made clear in the remainder of this
report, we have found that inaction andindeed, deliberate
inhibition of action, is a major factor blocking progress in
information handling activities of the intelligence community,
these expressions of confidence in the subpanel'n previous
activities are gratefully acknowledged.
Moreover, the portions of the report by the PSAC
ad hoc panel on non-numerical information processing quoted In
COMB 1963 ascribe to that panel a certain conservatism:
"The two fundamental difficulties faced in present
applications are that even though large-scale commitments to
the development of operational hardware systems have been made,
there is (a) a lack of precise definition and specification
both of the objectives to be reached and of the relevant non-
numerical techniques and (b) a shortage of first-rate research
workers in the fie1d.4
"Symptomatic of these ditttculties are th present and
growing external pressures towards large-scale hardware procure-
ment in cases where the operational purpose to be served is
vague and where, in any event, the present state of the art in
hardware and software is inadequate to support the grandiose
'intellectual processing' that is so ardently desired."
The initial requir ments of biographies are much more
Se_
modest. Our major finding B) that is is the so-called
Approved For Release 2002/05/09 : CIA-RDP80B01139A000500050001-3
4
) t
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
?ALT SECRET Page 12
a 0 CONTROLLEDD SSEM
ies" of organizational politici not any basic technical
obstacles, that spawn impossibilities in the biographies area &Aka-Ivo--
should not, therefore, be ascribed to any starry-eyed belief
by this subpanel in the miracles of automation. We think
rather, that no degree of automation can compensate for bad
management.
C. Definition of Biographies
Various intelligence community and related agencies
of the government maintain extensive biographical files to
serve the needs of Positive Intelligence, Security and,Counter-
intelligence an the case may be.
Our definition of these three major areas is taken
from CODIB V, Annex 1:
"C
const tu e
security.
of foreign
officials
activities
GENCE flIOGRAPHI
t activity
n on personal ties who
a known or pos5ible threat of national
normally include members and agents
intelligence services, Communist Party
and others engaged in organized subversive
GENCE BIOGRAPHIC AREA: That activity
noa onon persOilities, usually
o are of general interest to the intelligence
y. These include leaders in the scientific,
1, governmental, economic military, and other
p ssional/governmental fields.
"SECURITY BIOGRAPHIC AREA: That activity which deals
With Information held tirthose organizations which
have the normal function of investigating and granting
clearances on individuals or counterintelligence
interest in respect to the internal operations of the
holding organization."
1.?
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05/09 : CIA-RDP801811EAel5R0E0T3
DRAFT
SIICRE_____Z__FORN CONTROLLED DISSEM Page 13
Biographic files
include the CIA's Special Re
National Security Agency's ft
n the Intelligence Community
and Biographic Register, the
in the Office of Central
Reference,and files in the Defense Intelligence Agency and the
State Department. Major biographic files are kept' also by
the FBI, the Immigration and Naturalization Service, the Civil
Service Commission, the National Security Agency's Office of
Security, the Air Force Office of Special Investigation, the
Navy, the Army, and the CIA's Office of Security and RID.
Some consolidation of the indexes, at least, to certain of the
Armed Services files is evident at Fort Holabird. Additional
details on the nature and use of biographic files are given in
Appendices 1 and 3. Still greater detail is available in
CODIB V and SCIPS 1963.
D. Significance of Biographies
The importance of biographic tiles in Counter-
intelligence and Security is obvious since these activities are
concerned primarily with people. In the Positive Intelligence
areas, the biographic files are said to play a critical role
in support of the evaluation and production of foreign
intelligence but we have no erified this nor, apparently,
has the CODIB Task Team 5 (see Section II, A). In either Impo,
the collection, storage and dissemination of biographic
material is presumably not an end in itself.
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05/09 ? CIA-RDP806043?0t0R0E1-17
Dian
AECRETAWORN/CONTROLLED DIS34 Page 14
At present, the differentlocatorm and the real or
imagined differences in the uses of biographic flies in various
branches of the community lend existing files a considerable
measure of apparent variety.
A case can be made, however, that similarities on the
whole outweigh differences (Appendix 5). The members of
Task Team 5, hobbled by their terms of reference and apparently
hypnotized by the rigidities of the 80-column punched card
officially see incompatibilities and impossibilities wherever
two members disagree over the number of card columns that
should be allocated to such data as names or birthdates
the opinion of this subpanel the technical problems of
biographies are sufficiently simple in contemporary terms to
suggest biographic files as a primary candidate for competent
and honest experiments designed to explore the relative
advantages and disadvantages of various positions on the
spectrum from complete isolation to total integration of tiles
(Appendix 5a p. 1). The criteria for evaluating these experi
ments should be based on the satisfaction of explicitly
formulated needs of the community of users s to both
substance and security, not on catering to the crotchets of
keepers of the files.
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05/09 : CIA-RDP8(30C344180R30E0T-3
N .1MOT
OFORN/COMOLL
Page 15
B. Technical Simplicity of Biographies
The biographies area lends itself particularly well
to a study that cuts across the whole of the intelligence
community, in relatively simple yet non-trivial terms. While
the problems of creating maintaining and using biographic
files are by no means negligible, In comparison with mo &t
cii-
other files used by the intelligence community pri?reat
deal cleaner, simpler and better understood (Appendices3 5).
The biographic files are a very good example of files that are
widespread throughout the community, used throughout the
community* serviced by the whole community and which, while
relatively simple in structure, present all of the problems of
file maintenance in a non-trivial wayllySimpler files such as
order of battle or installation files exist in various areas
but1 in general, these are based on a narrower input spectrum
than biographic files and serve a narrower area of the
community. Because they can usually be rigidly formatted and
organized and accessed through clear-cut keys such as geographic
coordinates, their mechanization is well within the state of the
art and therefore
accomplished.
We do not wish to understate the ultimate complexity
of biographic files. We emphasize, however* that major
advances over the present state of affairs can be made without
fear of technical barrier
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05/09 : CIA-RDP80g1t9AC0
DRAFT
RIT/NOFORN/CONTROLLED DISSEM
0050001-3
E
Page 16
The report of the Biographies Task Team states that
"no major name index In the intelligence community has yet been
fully automated. Therefore proof of success has not yet been
conclusively demonstrated" (CODIBV, p. 2). The parallelism
between the restrictive phrase "in the inte igen?. community"
occurring in the preceding quotation and in the Task Team's
statement of purpose (CODIB V, p. 1) may account for the absence
of reference (see also Section III E) to such files as that
operated by the National Driver Register in the Bureau of Public
Roads of the Department of Commerce, which this aubpanel visited
(Appendix 3). This file, now accounting for 860,000 driver's
licenses revoked in each of the 50 states, the District of
Columbia, Puerto Rico, the Canal Zone the Virgin Islands and
Guam for offenses involving drunkenne s or fatalities got no
OV
consideration by the Task Team under the headings "major,"
"fully automated" and "proof of success," or any other heading.
We regard it as mos significant that such a central
service can deal effective].' sovereign states a* 01-1"-A
other jurisdictions. The degree of trust in the operation and
of understanding of its non-deterministic aspects (see the
" may be 's" in the Matched Driver Record Reports of Appendix 3)
is apparently such that the practice of returning no response
whatever to the state if a name is not found is generally
accepted, with consequent savings in transmission and paper-
shuffling costs. Registrars of Motor Vehicles apparently
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05/09 : CIA-RDP80131012EA0A5
DRAM
_ssallATApipmcp.__LE OILED DISSIN
understand and bel
eve
5rt
Page 17
there is a probablmatch in the
Register they will hear about it within 24 hours.
Admittedly, the National Driver Register Piles must
deal only with Intentional camouflage by driver* in danger of
license revocation while the community's agencies must some-
times deal not only with that, but with illegible, fragmentary
data, aliases, names overheard on noisy channels, etc. In the
panel's view, these problems are real and serious, but in no
way preclude immediate automation of a substantial portion of
the community's biographic files.
In commenting on the problem of variations
"categories of identifying data reco
repo i the Task Team states tha
the
impossible to
develop rigid rules on what constitutes the minimum identifying
data required" and that "each agency, in recognizing these
problems and the nature of its own index, forms its own rules
regarding minimum identifying data for recording and the depth
of search according to the nature of the request (corn v, p. 8).
Our site visits and conversations with members of the Task Team
suggest that the following exegesis of these comments merits
consideration: "Impossible" means politically, not technically
impossible, the "realities" get In the way. "Impossible" perhaps
means impossible with punched cards. There is no technical
known to this subpanel
reason/why only "rigid rules" must be considered when flexible
rules might do and modern storage and logic technology permit
their use (Appendix 5). In short, "impossibility" is a CODTB
thingApproved For Release 2002/05/09 : CIA-RDP80B01139A000500050044
Approved For Release 2002/05/09 ? CIA-RDIS
DRAFT
_S EC1_21TZNPFORN/C ?KROLL= DISSEM
leOR5
I as
5T01-3
Page 18
The foregoing should not be taken to imply that this
subpanel thinks that all the problems of the intelligence
community have been solved by the Bureau of Public Roads or that
"not impossible" is synonymous with "easy." It is, however,
clearly implied that there is a need for candid and unfettered
reporting of facts and for scientific experimentation.
F. Biographies as a Laboratory for Technology and
Management
The Clifford memorandum of June 15, 1965, (CLIFFORD)
characterizes the problems of the Intelligence Community in
the following terms:
"The problems of the intelligence community in
connection with information access and retrieval
include, but are not restricted to, those common to
all who must maintain very large bodies of informa-
tion in accessible form. This is even true in the
handling of information from unclassified sources.
The importance of negative information, and of patterns
of information, requires that access to intelligence
information produce a completeness of response beyond
that which is expected from many large files of stored
information. Like statistics, intelligence cannot be
satisfied with the highly anecdotal, but requires that
all available items of information are allowed to con-
tribute their part of the final summary or other
intelligence product. (p. 3, No. 7)
"As a consequence of intelligence community
requirements for high recall, the mechanized and
automated means of access to many sorts of intelligence
files cannot be required to meet simultaneously rigid
requirements as to relevance. Accordingly for some
time to come the mode of gaining access to intelligence
information will be through combined machine-human
systems that will seek the machine retrieval of stored
intelligence information in order that its relevance
may be established by human examination. It is this
combined machine-human factor which generates systems
problems of great difficulty and dimensions.
p. 3, No. 8)
Approved For Release 2002/05/09 : CIA-RDP80601139A0005000500
Approved For Release 2002/05/09 : CIA-RDP80Q011_39A900500050001-3
SECRET
DRAFT
SECRETOFORN CONTROLLED DISSEM
Page 19
"In the area of experimental approaches to the
adaptation of machine processing to the storage and
retrieval of intelligence information, an encourag-
ing beginning has been made within the National
Security Agency where the Technical Information
Processing System (TIPS) study is presently under
way. This experiment, although on a limited scale
and confined to a selected number of organizational
units and information files within the National
Security Agency, is producing important lessons for
the achievement of a realistic system for the
interrogation of a computer by remote users requiring
access to a common information base." (p. 4, No. 11)
In addition, Recommendation No. 2 concerning the
Technical Information System (T/PS) project (CLIFFORD, Page
5)
specifically calls attention to the fact that "the capability
for extensive handling of the Russian biography problem should
be available in the community-wide system by the summer of
1966."
The successful operation of some quasimeohanized
biographic systems in CIA, NSA and the Bureau of Public Roads
demonstrates that many technical and security problems arising
with biographic files can be solved at least to some degree.
Our observations in Section III suggest that no insuperable
technical problems blook coordinated action by the community in
the biographies area once management problems are resolved.
In particular, biographic files are free of most of
the serious linguistic problems which bedevil attempts to
handle the finding and retrieval of the extensive textual
materials that form the bulk of the intelligence community's
files. The CODIB biographies task force implies in one of its
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
SECRET/WOMEN/CONTROLLED DISSEM Page 20
Appnitied ror Alease 2002/05/0g : CIA-RDP801301139A000500050001-3
recommendations (CODIB V, Page 5, Rem) ndation 3) that a
successful biographic system could form a good basis for
approaching these more difficult general retrieval problems
through "a coordinated program (which) should be developed
using EDP methods to provide machine indexes of the biblio-
graphic data processed by any organization in this field, so
that the personality information is accessible to a recipient
in machine form, with quick followup to the translated source."
The converse would be equally valid, that biographic indexes
could lead back to bibliographic material containing informa.
tion about installations organizations, etc., related to the
personalities. This subpanel agrees that coordinated biographic
files together, perhaps, with installation and organization
files, may well form one excellent basis for indexing other
holdings of the intelligence community since curiosity about an
individual, an organization or an installation is so frequently
a source of curiosity concerning the more detailed and extensive
material that might be in the files.
The subpanel takes exception only to the mode of
implementation proposed by the COMB task team, namely that
another task team or its successor be tasked to fdaow up on this
recommendation. The record shows that COMB, as presently
constituted, has been singularly ineffectual in getting action
on difficult problems. Task teams have operated within
severely constrained terms of reference conducive to a bizarre
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05/09 : CIA-
CROFOT0001-3
DRAr 'J Page 21
SECRET OFORN CONThOLLED DISSEM
frame of mind in which members felt constrained to go at their
task as CODIB, deliberately avoiding the use of information
G-AE
available to them as individuals" (Statement to fulyanel at
Briefing of 4/19/66). Serious problems are recognized even
within these limitations, statements that strong efforts are
required appear, but the process seems inevitably to conclude
with bland ineffectual statements about cooperation and
consensus devoid of concrete suggestions for action and enforce-
ment of agreements.
The history of CODIB Task Team 5 - Biographies yields
an illuminating example of that subtle magic. While CODIB
itself boldly recommended in February 1964 to "Develop a
Biographic Intelligence Processing Plan," (CODIB 1964, p, 21,
Recommendation 4f), the Task Team's terms of reference as of
January 1965 (CODIB V, Annex 6) and the January 1966 Draft
Report contain only a limp eviscerated shadow of this goal in
the statement that "the objective of this Team was to 'identify
means for improving the storage, retrieval and exchange of
information from the major name files and related data files in
the intelligence community'. (See also Section III, DO
It is equally curious that while it was recognized in
1963 that "the most significant problem in the present period
is that of organization," that "collective efforts within the
community will become increasingly important" (BOREL, 1963;
Attachment p. 2) and that "steps can be taken within the
Approved For Release 2002/05/09 : CIA-RDP80B01139A0005000500 1-3
Approved For Release 2002/05/09 : CIA-RDP8S0
DRAFT
UREPIAMBELco TRO
DISSEM
900
11.11
Page 22
present state of the art to ease some of the current problems"
CODIB, 1963, p. 10), such pieties are transmogrified by 1965
into the idea that "attention (and management support)" be
given "to improving the individual systems in each USIB agency,
with projected compatibility monitored by CODIB and the
PFIAB/OST Joint Guidance and Evaluation Panel through bri fings
demonstrations and discussions."
The findings of the Task TeamVI on Research and
Development that "security barriers are used quite effectively
to barricade against attempts by other agency personnel to
acquaint themselves with ongoing R & D efforts" (CODIB VI, p. 13)
and that "very few intelligence staffs permit realistic
evaluation of their practices or products and results of any
evaluation are usually tightly held" (COD'S VI, p. 31) coupled
with the absence of observable experiments that would support
the change of attitude between 1963 and 1965 suggest that this
change is not as down-to-earth and commonsensical as might
appear at first glance. As of this writing, the report of Task
Team VI has yet to be approved by CODIB!
The biographies task team observed that, in the
positive intelligence area, "every organization has its own
standards for selection (from collected material) based on the
mission it is supporting and budgetary limitations" and that
"the same source document is frequently processed by different
PI organizations." They further remark that "there is an
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05/09 : CIA-ROPTE30CROE51-61-3
DRAFT
CONTROLLED D SUM Page 23
overlap of information in PI files because the different file
systems support the same requirements, or because the
personality mentioned in the source meets the selection
criteria for two different requirements. (CODIB V, pp. 13-14).
Given, on the one hand, that repeated handling of the Same
document is prima facie wasteful while, on the other hand,
interpretation of the same material from different and possibly
conflicting points of view is likely to be helpful and perhaps
vital to reliable intelligence production, it is sad but not
surprising that the task team felt unmoved or unable to recom-
mend any approach to this vital problem of management control,
especially since the same problems had already been recognized
earlier by the SCIPS Study (SCIPS, 1963).
Indeed, a disturbing proportion of the technical
people this subpanel came into contact with seemed conscious
only of details of card formats, while the extent of some
management personnel's appreciation of information technology
seems typified by such a curious anachronistic statement as
"our problems have largely to do with the processing of
language, while automatic equipment is essentially designed to
handle numbers" (BORK:, 1963; Attachment 1, p. 2). There is a
vacuum of competent consideration of system design problems
combining the understanding of needs and of possibilities.
This problem is not altogether unrecognized, as
shown by the following expression of the outlook for CODIB:
Approved For Release 2002/05/09 : CIA-RDP80B01139A0005000
Approved For Release 2002/05/09 : CIA-RDP80
5140t001r3
RET/NOPORN/CONTROLLED D18 SEN Page 24
Ne wifl need to guard against the ADP tail wagging
the information processing policy dog. But we do
need to give more emphasis to technical problems than
we have in the past." (BOREL, 1966, p. 6).
The problem is, perhaps, that the tail and the dog
have not yet met each other. As for wagging, while we have
repeatedly emphasized the need for keeping the user in mind,
we also believe that in a community where information is the
most important product, the processing of this information
should be far more than a managerial afterthought. ADP, as we
conceive it, has more to do with the head t/han he tail but
i
if pressed, we would insist at least on thinking kangaroo
rather than dog.
Where information is concerned, process and substance
assume nearly equal weights and management cannot safely assume
that it it pays attention exclusively to substance the
technicalities of processing will take care of themselves. The
top and middle managements of most organizations, not only in
the intelligence community, were trained in days when informa-
tion processing was for accountants or scholars, but not for
them. The fact that modern information technology affects the
very fabric of organizations has achieved some notoriety in
The New Yorker cartoons, but has yet to be translated into
realistic organization charts. (See also Section III, E).
It is the opinion of this subpanel that the biographies
,
area provides a unique raboratory in which to explore the
technical and management control problems raised by the Cl
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05/09 :
DRAFT CIA-RD?OE)1o9N011010001-3
aiN0 RN/CONTROLLED DISSEN Page 25
memorandum and to find out which of CODIB's several minds
reflects reality. The technical problem of biographies,
while by no means trivial, are not so
primary organizational and management
COINS effort may be a beginning.
Any experiment, however we3.
worthless if set up in the CODIB pat
are evident from the following expres
severe as to becloud the
control problems. The
intentioned, will be
n. The reasons why
on of the COMB credo:
Ne recognize and accept certain constraints upon
what can be accomplished by joint action because of:
the primary responsibility of USIB members to
their command channels for carrying out basic
departmental and service misSions; and,
the impact of the NSC allocation of intelli-
gence collection and production responsibilities
among agencies upon supporting information
processing programs." (BOREL, 1966, p. 3)
The deadening effect of the imposition of such
administrative constraints in made explicit in the recognition
of the following weakness of CODIB and its approach:
"Common denominator of agreement may be so
large as to negate utility of the solution."
(BOREL, 1966, P. 5).
That these constraints bring out the worst or
ubiquitous bureaucratic tendencies is made plain by the less
elegant advice one of our informants assured us applies under
the circumstances: "Cover your ass." We can't help thinking
that this motto is as inappropriate a slogan for a healthy
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05aMA-RDFSBE1
OrOWCONTROWED DWEM
Ii .
E5101-3
Page 26
intelligence community as "Expose your rear would be for a
crack military unit.
We have, in addition, fair evidence that "the system"
not the people is inherently responsible for this sad state.
Individuals whose wits seem hopelessly dim in an official light
visibly brighten in the sunshine of informality. This
phenomenon is common to all organizations, but seems pathologi-
cally pronounced in the Community. While, once again, we
recognize the need for users and operators to exert the
strongest influence on systems design, we recommend that
members of any group constituted for this purpose be instructed
to act according to their individual wisdom and conscience and
that they be protected by appropriate administrative devices
against any wrath they may draw from ho as a consequence.
Someone, presented with conflicting opin one instead of bland,
lowest common denominator consensus may then have to make
wre
painful decisions, but we submit that the Communit
A
management able and willing to act in this way.
We do not think his recommendation is naive and
inapplicable. We found it refreshing to learn, at Fort
Holabird, that the very same people who saw only impossi-
bilities, parochial interests and lowest common denominators
when acting in a COD1B group, stopped dragging their feet and
leaped into action when a directive issued by the Secretary of
Defense on May 27, 1965, ordered them to provide a centralized
;
Approved For Release 2002/05/09 : CIA-RDP80B01139A00050005000 -3 E
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
DRAFT
SECRET
0 ?age 27
( Sea_ aQ4c ScAlis,5)
index to the Armed Services investigative files by May 27, 196
Indeed, the job apparently got done. Critics do point out that
excessive haste has, in their opinion, produced a product lack-
ing elegance, generality, or power. Nevertheless, the illustra-
tion shows that management has a real choice. Under the
constraints described in its credo, eau had none.
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05/guice-RDP8OB 3Or0R50E1t
Page 28
SE RET/NOPORN/CONTROLLE D DISSEM
OBSERVATIONS
A. Relative Characteristics of Positive Intelligence,
Counterintelligence and Security Files
The Positive Intelligence files are in a sense the
least seneitive of all. Since they deal largely with foreign
personalities and a substantial proportion of their source
material is from the open literature, the problems of indexing
these files are subject only to the normal technical constraints
common in the Intelligence Community, without touching more
delicate areas such as might be raised in the evaluation of
uninterpreted data on U. S. citizens - where the very mention
of a name in the file can create a problem. On the other
hand, the files are among the more problematic in terms of
acquisition, indexing and completeness.
In the Security and Counterintelligence files a
goodly portion of the inputs is on controlled forms often
supplied by the individual himself and positive identification
In tags, such as Social Security number, etc., are readily
possible. The PI files, on the other hand, are constituted
from a much greater variety of sources they tend to be far
more fragmentary and their use is more difficult.
It therefore seems reasonable to infer that general
volutions of Positive Intelligence biographic file problems
could be readily specialized and adapted to the needs of
Security and Counterintelligence file but that the reverse
;
Approved For Release 2002/05/09 : CIA-RDP80B01139A00050005000
Approved For Release 2002/05/09 : CIA-RE1060E1C9PROrT001-3
DRA7T 0
S RET/NOPPN/CONTROLLED DISSRM
very likely would be far more difficult. These stat
Page 29
are
supported by evidence to the effect that PI requests tend to
include 20% name finding, and 80% name searchini while in
the Counterintelligence and Security files the proportions are,
respectively, 5% and 95%.
The study of PI files thus presents the greater
challenge to research and development as well as a hope of
finding general solutions. Such study should be encouraged.
In the near future, however, useful practical results are
much more likely to be obtained by concentrating on the auto-
mation of CI and Security files.
B. File Overlap and the Single File Problem
CODIB V points out (page 28 paragraph 6) that "there
are several thousand people involved in biographic activity in
the Intelligence Community, Approximately 1000 of these, at
an annual salary-only cost of $5,000,000 are directly involvedert,..
work at the index level, in the preparation, maintenance and
searching of the major biographic indexes. The report then
goes on to point out (page 13 2.b) that
The terms "name finding" and "name rehing defined in
the co= Task Team V Report as follows:
"NAME FINDING: The searching for name information about one
or a group of individuals by looking for data elements other
than the name, such as date of birth, position, location,
organizational affiliation, occupation, military rank,
nationality, including a combination of such factors."
"NAME SEARCHING: Search of indexes or files organized by the
names of persons to determine if information exists on the
Individual, or to validate basic information."
CODIB V, Annex 1, p. 1, Glossary.
Approved For Release 2002/05/09 : CIA-RDP80B01139A00050000009
Approved For Release 2002/01RAJA-RDPESBOF3r0ORT0131-3
SECRET/NOFORWONTROLLED DISSEM Page 30
"The basic criterion of any agency for selecting an
item for a PI file is whether or not the item supports
the foreign intelligence effort on a particular country
or area. Every organization has its own standards for
selection based on the mission it is supporting and
budgetary limitations. The same source document is
frequently processed by different PI organizations ...
There is an overlap of information in PI files because
the different file syatems support the same require-
ments, or because the personality mentioned in the
source meets the selection criteria for two different
requirements: e.g., CIA and State have an interest in
military personalities who are prominent in other fields
such as politico, science, space, etc., whereas DIA and
NSA are interested in the same person because he is in
the military field. There is no assurance, however,
that because a personality is mentioned in a source
document that he will eventually be processed into a
PI file."
Further on, the same report points out (p. 152 4.c) that
"Some files are restricted by security classification
as to what can be processed. Research in such a
limited source file often gives incomplete or out-
dated information."
The CODIB report further states that
"It is doubtful that any single file, whether it be
computerized or manual, can ever be considered a
complete or sole source for biographic information."
This subpanel takes strong exception to this last statement
largely on the ground that it has not been able to uncover
the existence of any data, produced as a
Q..s5v,c9--c-LA2,chr
hensive systems stud:yAmIkthe advantages
si le community-wide biographic fil
result of a compre
and d sadvantages of
SC( G7-:? )
It should also be
noted that the Clifford memorandum (CLIFFORD, Recommendation #2)
specifically suggests the Russian biogrphia problem as an
experimental vehicle for the expanded TIPS system (COINS).
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05/DRAM-RDP8S0E9e0
SECRET/NOFORN/CONTROLLED pissEm Page 31
tto1
?ET-3
The evidence available to us suggests that consolida-
tion or at least common indexing and easy mutual accessibility
of CI and Security files is technically feasible. The issue
of whether or not this should be done was left in limbo by
CODIB's departure from its resolution to develop a biographic
intelligence processing plan" (see Section II, F).
In the PI area the technical issues are more confused
and therefore all the more in need of candid intelligent
Investigation. For example he prevalent practice of scanning
PI input materials for names, and resolving name variants and
name variations before entering the material in files has led
not only to a substantial backlog of unentered and therefore
unavailable materill but also to the repeated resolution of
the same problem by every recipient of the same material.
The alternative of entering materiels only once
for the whole community, using as a key whatever name variant
or variation appears in the raw material, and applying name-
grouping techniques to draw together scattered relevant
material at the time an actual request is made for informa-
tion has received only scant attention. Such a task is
t;.?re,
clearly with manual or older punched-card techniques,
but not at all ruled out with modern equipment. The feasi-
bility, potential dollar lavings and increased information
availability that such techniques might provide must be
investigated in spite of the specter of centralization
(Appendix 5).
-
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05%:p9A-RDP80136 ACO5
SECRET/NOPORNJONTROLZED DISSEM
5E0113
Page 32
The SCIPS report points out that, although "DCID 1/9
allocates responsibility for production of biographic intel-
ligence and the collection and maintenance of biographic data
on foreign personalities," (SCIPS 1963, p, 57), various
agencies (NSA is singled out), maintain substantial biographic
1.141
files without specific ?cation in DCID.41604(lso, as 4111PW
0-side neidenae of many files that are no longer input
)
tel."
The report then comments that "a set of huge files
with their implied amount of processing effort is being main-
tained in a manner not identifiable in terms of DCID func-
tional responsibilities. This is not to say that NSA does
not have a need for this information, for it manifestly does.
Therefore the inference is that those responsible under DCID
are not performing satisfactorily to NSA needs or that subject
category is not a good way to allocate this responsibility or
both" (SCIPS 1963, p. 68)
Had CODIB undertaken "to develop a biographic
intelligence processing plan" some light might now shins on
this obscurity. They did not and therefore it does not.
C.. Evidence of Current Automation Efforts
There is encouraging evidence of widespread begin-
nings of modest automation efforts throughout the Intelligence
Community. Within the CIA the Special Register, the Biographic
Register and the Walnut system
oeiated indexes are
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05/09 : CIA-RDFSBE1COR5CE09001-3
DRAFT
/No Hie? Page 33 OLLED DISSEM
mechanized to varying degrees and with a variety of different
technical and file structure approaches However, in some of
these cases, punched cards were introduCed 10 years ago and
nothing new has happened since. Wholly or partially mechanized
biographic files exist also in the NSA, DIA, and at Ft. Holabird.
However, as discussed above (Section II, F) community
effortsto mechanize biographic files are fragmented and
generally uncoordinated between agencies. In view of the
recognized overlap of biographic files (CODIB V, p. 14, pare 20),
it is difficult to understand why a community-wide effort to
evaluate the possible support of ADP has not been sponsored,
except on the hypothesio that management has been unwilling to
sponsor it. Why CODIB oes not recommend such action is
clear from its charter charter and the pressure toward
the lowest common denominator. The irony of this situation
is made plain by Recommendation 8 of CODIB V which states (CODIB V, p.)
"The Task Team III (or its successor) be tasked to
study those various programs exploiting open source
scientific and technical information which generate
personality information of positive intelligence
value as a by-product. A coordinated program should
be developed using EDP methods to provide machine
indexes of the bibliographic data processed by any
organization in this field, so that the personality
information is accessible to a recipient in machine
form, with quick follow-up to the translated source."
(See also Section II, F.)
Apparently, community-wide mechanization is good in this
bibliographic area but not good in the derived biographic
area. This subpanel has not perceived any substantive
Approved For Release 2002/05/09 : CIA-RDP80B01139A000500060
Approved For Release 2002/0WKIA-RDAE &0650 01-3
SECFM/NOPORN/CONTROLLED DIMES
Page 34
reason for this distinction, and our recommendations reflect
this fact.
D. Need for Research and Development
There is some encouraging evidence of recognition
of the need for research and development activities in the
biographic* area. Notably, the COMB V Report stresses the
critical technical problem of dealing with name variants and
name variations in name searching through biographic files
especially those with a preponderanceof foreign names or
even of U. S. citizens in cases where
is not possible (see also Section III
Problems created by the high proportion of common
names in various files (CODIB V, p. 8 item 8) are also
recognized, as is the _more difficult technical matter of
providing usefully formatted identifying data on named or
unnamed persons to permit searches of the name-finding type
or identification in the absence of names.
Nevertheless, the Panel is encouraged by the finding
of the CODIB Task Group (CODIB V1 p. 11, item 5) that "in
observing some of these typical and widely used forms, the
Team found that certain basic data, such as name, place and
date of birth, Service Serial number, Social Security number,
sex, etc., were included on each form." While this finding
applies primarily to the Cl/Security files, perusal of the
Chart in Annex 4 of the CODIB report indicates that an effort
source format control
Approved For Release 2002/05/09 : CIA-RDP80B01139A000500050a01-4-?
Approved For Release 2002/05/09 : CIA-RDPAQW,1132,A000500050001-3
DRAFT st,GRET
SECRET/NOFORNAONTROLLED DISSEM Page 35
is being made to provide similar fields, even in the Positive
Intelligence filet although the probability of no information
appearing in these fields is obviously greater. The fact that
certain standard elements, blanks in the formsyls it were
EJL
may occasionally be left empty inCalsmishoubmwmforms does not
negate the clear possibility of defining standard elements
even in PI files. The specter of empty fields haunts those
accustomed only to punched cards, but is readily exorcised by
more modern techniques.
As pointed out in the Introduction to this report,
the principal technical issue faced by the intelligence
community is that of making maximum use of the technology and
methodology available in the nation, In some problem areas,
there will be a need for directing resources toward the
development of technical solutions designed especially for
biographic information processing problems. In order to guide
such efforts the community will need a coordinated R & D
program and strong in-house capabilities in specific areas
which are not covered in other government or noagovernment
agencies.
(The original CODIB Terms of Reference (DCII) 1/4)
charge CODIB with the responsibility and authority to develop
a long range plan for R & D in information processing. HOLD
FOR CHECK OF 4/23/65 TERMS AGAINST 6/6/9J) CODIB has never
A(4K_.0? ru
4-ttA /):4 r HA Y
Approved For Release 2002/05/09 : CIA-RDP806011641a04 00 12V
Approved For Release 2002/05/444A-RDP80B011SAEO5COrre
E.' T
ECRET/NOFORIVCONTRObLED DISSEM Page 36
concerted action in developing such a plan; the torte of
Task Team VI represent an admirable start but its tartly
critical report has been referred for further drafting by
41110 usual CODIB technique. The effect of CODIB action in the
R & D area has been to prevent other groups from initiating
a community-wide program by keeping the R & D responsibility
preventing itself
C 10810WOMmap in its charter and, by inaction,
from initiati such a progrem.+
The biographies area is merely one of many that are
hampered by this policy. It is therefore understandable that
CODIB V did not recommend any R & D activity as an aid in
solving biographies problems; the Task Team objective appears
carefully emasculated in order to preclude any consideration
of fundamental methods for solving biographic problems.
The Task Team was barely able even to state the
roblem under its modified objective. Finding #1 (CODIB V,
1) states that:
"1. Improvements in the speed and quality of biographic
information processing involving interagency exchange
on U. S. citizens and foreign nationals are necessary
to further improve security, and to afford policy makers
and analysts better response from biographic intelligence
files on foreign nationals of interest from a variety
of angles - military, subversive, political and scientifte.
The Team finds that use of computer techniques and inter.
agency telecommunications links may provide significant
improvements."
nding #2 then remarks:
"2. There are, however, profound, complex problems and
significant costs in making major changes in the large
biographic holdings of community concern, particularly
if the changes involve conversion to computer systems."
14
Approved For Release 2002/05/09 : CIA-RDP80B01139A00050009-100111,44"
Approved For Release 2002/0%1 ? 1A-RDP80,001_1_39A000500050001-3
SECRET
AmFoRIVONTRoLLED pissm Page 37
The inclusion of policy makers in Finding #1 implicitly argues
for the importance of the files and the associated problems.
The recognition of the profound nature of the problems in
Finding #2 implicitly argues for the need for very strong and
powerful recommendations for action in the R & D area as well
as other areas.
The actual recommendations made by the Task Team are
pathetically inadequate when held up against the problem as
stated in Findings #1 and #2 However, when the recommenda
tions are compared with the Task Team objective, the fit is
exceedingly precise. This subpanel can only conclude that the
original objective was modified (see Section II, F) in order
to place strict control on the recommendations of the Task
Team. TcflubPanel suggests that in the future control be
placed on the findings rather than on the recommendations;
this would seem to be a more effective technique for the
apparent purpose.
Given the context s eubpanel understands why the
Task Team could not make recommendations in the R& D area;
therefore this report includes R & D recommendations the
Task Team should have made (Sections I C; III 8). Faced
with Findings #1 and #2, this subpanel would be remiss to do
less,
E. The Nature of the Biographies R & D Problem
There are many biographies problems that can be
solved with available technology. However, there is a wide
Approved For Release 2002/05/09 : CIA-RDP80B01139A0005000S30i3
ta 0 -T
Approved For Release 2002/05/09 ? CIA-RDP80 1 9FACD5 0 3
DitAn
SECRET/NOPORN/bONTROLLED DISSEM Page 38
gap between available technology and the community's capa-
bility to apply it, It is apparent to the subpanel that a
new qualification, namely computer systems analysis, must be
added to the community's wide range of expertise. This
qualification must be distributed broadly and at every level
including the highest. Once a major fraction of community
personnel have some training and experience in this discipline,
it will become much easier for the community to avail itself
immediately and rationally of the benefits of off the-shelf
technology and methodology.
Ancillary benefits will also come in the R & D area.
Specific and important benefits should accrue directly in
biographic problem areas. The most immediate and important
benefit would be an appreciation by senior management personnel
responsible for biographic file installations and their use of
the need for and the nature of R & D work. With appreciation
and understanding would come stronger leadership and better
control.
Of the entire listed membership of Task Teem V, no
single person has the experience or qualifications of a pro
fessional EDP systems or applications analyst. A few have
medium heavy experience, but the rest have very little1 if
=cc]
any, experience in applying EDP equipment oblems.
This casts no aspersions on the personnel involved, because
there tvA3 not been a need in the past for such experience.
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/04D?TCIA-RDP801301139A000500050001-3
C E 'Page 39
SECRET/NOPORN/CONTROLLEP DIESER
Most biographic files were started as manual systems; the need
for mechanization is relatively recent. Thus the lack of
highly qualified personnel in biographic file systems may be
seen as the root of the lack of appreciation for R & D work.
But even granted the existence of a cadre of highly
qualified EDP systems analysts in biographies work, there
remain the "profound, complex problems" of Finding #2, and
these have to be solved,/these problemsx combining the
technical and the administrative therefore problems for
which methodologies or technologies cannot merely be picked
off the shelf.
In order to shed further light on these problems it
is instructive to examine the Task Team Findings (Section III, D)
and their atatement of the problems from the viewpoint of a
professional EDP systems analyst.
Finding 01 begins with the clause "Improvements in
the peed and qu-ility of biographic information processing
are necessary ..." This statement i very good because it
emphasizes speed and 9uality, two of the three basic parameters
by which any information processing installation can be
measured (the third parameter is cost). The Team further found
"that use of computer techniques and interagency telecommunica-
tions links par provide significant improvements." At this
'Our emphasiS.
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05/01934&-RDP80B21
AMORVOONTROLLED DISSEM
or
Page 40
point, the Team reveals its level t expertise; thei is no
real doubt that computers and telecommunications links can
improve biographic information processing because of the high
capacities (bandwidths) of both modern computers and tele-
communications equipment!
Evidence if needed, is f und in the fact that some
members of the community have already installed both types of
equipment (e.g., CIA-Walnut and LDX) precisely on the ground
that both quality and speed are thereby improved. The best
evidence for the feasibility of improvement via EDP is there
fere found within the community itself. A professional
systems analyst would also be able to deduce this trivial
conclusion from an examination of the files themselves. For
the moot part they are formatted, and even the unformatted
files (such as in CIA-BR) offer no technical impossibilities
(Appendix 5)G6st is the most important determinant, barring
politics, of course.
Finding 0 ie somewhat ore instructive. Here the
emphasis is on "making major changes particularly if the
changes involve conversion to computer systems. This state?
ment distorts the realities of conversion (Appendix 5c).
Usually, the cost of the job is the only hindrance and it is
hard to believe that any other parameter could be seriously
considered. The "profound, complex problems" have been
solved in several community installations visited by the
Approved For Release 2002/05/09 : CIA-RDP80B01139A0005600500M-3
Approved For Release 2002/05/09 LcIA-RD,E8Off1 trirry001-3
DRAFT 0 C.,
ET/M0FORWOONTROILED DIPBEM Page 41
subpanel; usual y the profound complex problem turned out to
be changing the attitude of management. This change, however
can hardly be charged as a technical component of the systems
problem.
More to the point it should be noticed that
Finding #2 sees the problema as consequences of making *hen ea
in the file installation. In other words, the new file will
be a modified version of the old file, and the greater the
changes, the greater will be the problems introduced. The
inevitable conclusion is that the best way to solve the
problems is to make no c nges at all. This gratuitous finding
thus gives the couip-de-grace to Finding #1 which weakly states
that changes k'ir;TrievA may improve the speed and quality of
biographic information processing. The professional systems
analyst would not be intimidated by Finding #2, thus leaving
the subpsnel with the conclusion that this Finding was intended
for readers who are not systems analysts.
The remaining Findings offer b.tantil data for
the systems analyst.
Findings #3 through #10 presentn numerical
data that is very useful in dertV ng a picture of the biographic
file problems. Indeed, the problems are great, but by no means
beyond the limits of technical possibility.
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05/
PORN CONOLLED DIS
C' r.
RDP80B011SA600510e040r3
Page 42
Finding #8 is worth a modicum of attention. Despite
the Implied threat of Finding 0 Finding #8 says that for
name searching, EDP may provide Improvements, but this state
ment was apparently considered rash and puerile, for it was
immediately slugged in the following sentences with the pompous
and nuous y true pronouncement that no major name index ..
has yet been fully automated, (see also Section I, E) thus
invoking the threat of Finding #2
In Finding #101 the body of FInding #1 is still
twitching. We are told here that name finding may possibly
be improved, but only by exchanging information about "EDP
techniques for improving speed and flexibility of response."
Notice how the golden "quality" of Finding #1 has been trans-
muted to leaden "flexibilit
Finding #11 then concludes with a wideeyed view of
the wonderful world of professional interchange. The whole
net of Findings reminds one of Franck 's Symphony in D where
the good theme jousts with the bad theme. In CODIB V, the
bad theme wins, and Finding #11 depicts the rainbow after the
storm.
The foregoing considerations while necessary to
achieve catharsis, do not take the most constructive viewpoint
of the problems. The standard approach of systems analysis
Is to create something new out of something old: change is
not something to fear, but something to use to advantage.
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/0401g1A-RDP8OBS1EA0e5ft5eilf
sEcREr1opoRNICONTROLLED DISSEM
Page 43
PreoccuP*t ion with format standards can prevent one from
seeing a better solution to the overall problem. Annex 4
(CODIB 17) illustrates the reality of the problem in different
agencies. There seems to be little doubt that each agency will
need more or less data in its records than another agency.
There is no technical reason why each agency cannot have
information entered, stored, and retrieved in any format of
its choice. Annex 4 merely tabulates the differences which
the system must accommodate.
From a systems design viewpoint t would be a
mistake to try to force standard formats on every agency when
there are more significant and difficult problems to be solved.
As discussed earlier in this report (Section II ), bio-
op.?41 v4.)
graphics offers unique rtwttie3 as a vehicle for R & D.
Sample suggestions are as follows:
?????
1. What would a single communitywide biographic file
installation be like? Specify its size in storage
capacity, response speed, request rate, etc. What
administrative problems would it create? How serious
are these?
2. Develop methods for recognizing proper names in
unformatted text and for extracting text which includes
references to or data about people.
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/0Mi+IA-RDPEEVE3001940?541-3
AOFORN/CONTROLLED DISSEM
3Develop a specif
a
Page 44
on for a very low costtech-
nology for entering data into tiles, to be installed
1972-75. Compare with cost projections of optical
character recognition technology. Will the latter be
the final answer? If not, a different approach must be
developed and work should be started now.
cifications for
nology that
technology1970's. What type
h perform t?
F. The Invisible Colleges
There is a strong need for positive interagency
cooperation, information exchange and joint enterprise in the
biographies realm. This need is clearly demonstrated by our
observation of efforts in several agencies without evidence
of mutual knowledge. This impression, which might otherwise
be attributed to the superficial character of an inquiryyby
outsiders, is corroborated by Finding 011 of the CODIB Task
Force, which states;
"The Teem agreed that the professional interchange
derived from the Task Force effort was highly valuable
to each member in providing new insights in manual and
machine techniques, interagency channels, sources of
information and policies of bther agencies."
A similar reaction was observed by the Panel in the o of
a meeting on March 29th 1966 which confronted representatives
of the Electronic Data Processing arms of NSA, DIA and CIA,
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05/911A94-RDP8OBOS3
ECRET/NOMINMONTROLUD DISSEM
torri
Page 45
with representative. of the counterintelligence offices in a
discussion of security and compartmentation problems. Om
l'nvisible colleges i.eI.* responsible for the work that does
get dor 9ot adequate .12 cikkAZNI1 11 -re Q-C,-,r; ???-??-?.s cha C
G. Compartmentation and Security
There is convincing evidence that eompartmtatjon
exists not only for security reasons but as the R & D report
points out (COD VI, p. 13) also for administrative reasons
masquerading as
"security barriers that are used quite effectiv ly to
barricade against attempts by other agencies' personnel
to acquaint themselves with ongoing R & D effort."
The problem of file security fails to be adequately distinguished
from administrative and in-fighting problems. We acknowledge
the genuine needs for security and compartmentation but we
believe that these have been grossly exaggerated wlith adverse
effects on Community cooperation.
The Community clearly recognizes the possible damaging
effects of compartmentation witness the statement
"Research in sueh a limited source file often gives
Incomplete or out-dated information." (CODIB V, p. 6)
Anecdotal evidence of the effects of delayed or
Impossible collation of relevatit facts abounds, but we notice
a total absence of either a mechanism for studying this problem
or even a genuine concern for it. The OODIB Biographies Task
Force report nimbly skirts around the problem by making such
Innocuous recommendations (CODIB V, p. 4, item 5) as that
Approved For Release 2002/05/09 : CIA-RDP80B01139A0005ak OP00-
Approved For Release 2002/05/W4A-RDP806011SA0E5C05RO1
/NOFORVOONTROLLED DISSEM Page 46
"the CODIB Support Staff be directed to prepare and maintain
current publications to inform users of biographic information ..."
with the hedge that this effort be limited
"within the limits of security classification and need-
to-know prescribed by each attenCY**"
This recommendation firmly endorses motherhood with zero real
effect; in fact, it flatly contradicts the one previously
quoted from page 16 of CODIB V.
H. Conceptual and Managerial Problems
The R & D report rightfully points out that
"Conceptual and managerial problems are more crucial
than purely technical." (CODIB VI, P. 3).
The R & D report further points out (CODIB VI p. 5)
that
4There is neither an organized set of R & D objectives,
a policy for establishing R & D objectives, nor a
mechanism for accomplishing either."
It is clear to the Panel that the product of one
agency or compartment is the input of another; that there are
many loops, cycles, interconnections, etc., but that the
prevalent concern for local optimization of resources and
performance is mailighat by any visible concern for total
systems design.
There is an almost total absence of planning data
or statistics in a form that would enable management to
determine appropriate courses of action. This Panel would
Our emp
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05gARA-RDP8013 U
,0)10A941351211011QQ0,I4?
-.)r.11t.
SECRETATOFORN/CONTR9LLED DISSEM Page 47
welcome a refutation of this statement in the form of appro-
priate data. For instance although the question of security
of shared files is agreed on by everyone to be of primary
importance, no one can supply useful data or criteria concerning
such elementary matters as the relationahip between the level
of security precautions and expense and the safety acquired at
such a level. It is striking in fact, that the Chairman of
the USIBSecux'ity Committee regarded this question as a nove
one when it was asked during the Security Briefing of the Panel
on March 20, 1966. The data may exist somewhere, but the fact
that this Panel was unable to uncover it is itself symptomatic.
Elementary facts concerning management techniques should be made
available if a decent system is to be developed. ,
We have dealt briefly elsewhere in thia report with
our concern that process become less of a stepchild (or tail)
of Substance, with the need for users to participate in systems
design, with the problems of enforced consemus and lowest
common denominator, with the need for a probabilisticview of
collection and retrieval, and with the need for introducing
competence and understanding of systems design problems at the
highest level of management.
At the root of much o
over the
introduction of automatic information processing technique
there is a fear, not only about the matter of security, but
also about agency or management authority. After allfolk
wisdom tells us that knowledge is power.
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05/09 : CIA-RDP80601139A0005
DRAFT
8ECRZT OPORN CONTROLLED DISSEM
IRET
Page 48
This is a valid concern, but we suggest that it can
be exaggerated, and that the fear is a large extent fear
of the unknown. While there are legitimate reasons to guard
information, at least part of this concern arises from a
mistaken confusion of information gathering with the exercise
of authority. Clearly, the opening of information lines up,
down and acre.ss would legitimize a leaping over organizational
boundaries that, while essential for real accomplishment, is
done nowadays only at official risk and peril. Organization
lines reflect lines of authority, but while knowledge is power,
the gathering of information is not the exercise of authority.
It seams, therefore, perfectly proper for a manager to leap
several levels down in search of answers, for a subordinate to
leap across organization lines and occasionally over his boss's
head, so long as decisions and orders travel by normal channels
and care is taken to protect legitimate privacy.
This vital question must be dealt with if real
progress is to be made in Community-wide information processing.
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
Approved For Release 2002/05/09 ? CIA-RDP80B0j1,39App05
DRAPT
;
050001-3
SECRET/HOPORN/CONTPOWD DISSEM Page 49
I. Manpower
The Subpan l believes that there is a serious
shortage of skilled manpower applicable to biographic prob-
lems. The skills in short supply include both computer sys-
tems specialists and biographic specialists. There appears
to be a wide variation in competence and experience of these
people in the installations visited, and the absence of poal-
tive communication channels precludes a sufficient interchange
of experience which could help upgrade weaker personnel.
The Subpanel is impressed with the quality of ex-
pertise found among senior biographic personnel. However,
these specialists appear to be made, not born, and their skill
Is the result of extensive on-the-Job experience. At the same
time, this pool of talent is fragmented along the same lines 83
are the files. This strikes the Subpanel as an unwarranted
division of a critical capability. To have experienced indexers
review identical source materials in several agencies and yet
have analysts be unable to have view of the total picture of an
,t17
-- individual strikes the Subpanel as a serious deficiency. We
applaud th intent of Recommendation #8 of the Biographies Task
Team Report which recommends in part,
"A coordinated program should be developed using
SDP methods to provide machine indexes of the
bibliographic data processed by any organization
in this field, so that the personality information
is accessible to a recipient in machine form, with
quick followup to the translated source."
'4()__15D'i S Ik f ; - er-%
Approved For Release 2002/05/09 : CIA-RDR8q13011 A 00500050001-3
Approved For Release 2002/05/09 : CIA-RDP80
W3A9Q
O;00050001-3
DRAFT
SECRET/NOFORN/C ONTROLLED DISSEM
-
..111116.0.
Page 50
tiailliarrairo----- Considering the strategic necessity of a pre-
eminent capability in biographies, the Subpanel feels that more
than good intentions are required.
Therefore, it is r commended that a plan for develop-
:a
Ing a community-wid captbI?Ity with the highest professional
standards of performance be submitted by 2 Jan 1967. This plan
AAA-kJ:L.-A-3
must includesmiNWPAMOMmtfor upgrading the skills of individuals
to a much higher average level and It must also include plans for
Wit
adding EDP expertise to the, important ap bility. Appendix 5
of this Report presents some specific suggestions which should
have been included in the Biographic Task Teem Report.
J. Communications
Recommendation 4'2 of the Biographies Task Team Report
says that each Agency should
Study the feasibility of telecommunication links
within the national agency check complex to facili-
tate the exchange of requests and replies."
In this connection, it should be noted that the extent of this
problem and its security implications may have been exaggerated.
The figures compiled by the Biographies Task Team reveal that an
extraordinarily high percentage of requests addressed to bio-
graphic files are returned with an indication that no record
), 4.)
could be four 4. This fraction varies ffom ;0% for Civil Service
Commission tiles to a low of 2% for DIA files. For positive
f.,LE3
Approved For Release 2002/05/09 : CIA-RDP8 1.11.19A00 500050001-3
Approved For Release 2002/05/09 ? CIA-RDP8OB
DRAFT
SECRET/NOFORN/CONTROLLED DISSEM
0050001-3
Page 51
intelligence requests, three of the five agencies cited stated
that 50%-of the responses were no hits. Since this response can
be communicated by one bit (or none) of information, it strikes
this that neither traffic density nor security plays a
ci-Ltt A C major role role and that current time delays are o
ndi There
oug?ckirv,z
is no evidence-eit ~ear
4
diversification and multiplicity of the files.
This Subpanel believes that telecommunication links are
obviously feasible and here is no need to study anything except
the costs and configuration desired. This is routine engineering
work and should be implemented within 6 months.
K. An Approach to Security Problems
We note with pleasure a tradition and an experiment
which suggest that the security problem for biographic files may
be tractable. The Biographic Task Team Report points out that
"Information about individuals comes from a great
diversity of sources, through a large number of
channels, and has been stored in a variety of
retrieval systems in diverse formats." wage 6, item 1)
4151)3-i3 V
f the no-hit responses are due to the
The Report then goes on to point out that
"These have traditionally taken the form of index
references, either self-contained or leading to
dossier files or individual documents."
Here, the tradition of the Community seems in complete accord
with the needs of a readily accessible, mechanized or unmechanized,
biographic file system. The CIA experience with Walnut and its
associated systems reinforces this impression. The Walnut File
contains at least two major types of biographic documents; one
Approved For Release 2002/05/09 : 00500050001-3
f,
sav,-cS
Approved For Release 2002/05/09 : CIA-RDP80B011-39A000500050001-3
DRAFT )
SECRET/NOFORN/CONTROLLED DISSEM Page 52
of the counterintelligence type, the other of the Russian
biographic type - to which access is controlled by different
compartments.
There isfebilialOWejea unified index to these collections
but security compartmentation is still maintained within the in-
dex as a safeguard. The separation of index and dossier leads
naturally into it strirtified retrieval technique which could
S4-0n6, valffile fras2A7pof security' lesiose- he pace of mechaniza-
tion. Index file volumes are considerably less than dossier
file volumes and the one can be mechanized while the other is
not. Together with the use of such techniques as LDX trans-
mission, this may lead to reasonable solutions.
The separation of index and dossier leads naturally to
the concept of free search on the index with a filter at the out-
put. This is, in fact, the way in which the Walnut File operates
Anyone who is authorized to use this file has access to the index,
but when a hit is made in the file,
is authorized to see are returned to
a control point and the requester is
permission of the controller to have
di
only items which the requester
ed to ha ye I
in such a sensitive
technique
him - the others are sent to
notified that he must seek
access to the data if he is
he fact that this kind of system can work
area is most encouraging. The fact that this
seems to be unique illustrates how slowly ideas travel,
even on innocuous matters of technique. COINS is presently trying
to grapple with some of the problems
1-k k
%eas2T05A P
of adding biographic files to
s,-54
VWS101002000
Approved For Release 2002/05/09 : CIA-RDP8017 1411500050001-3
DRAFT
Page 53
SECRET OPORN CO TROLLED DI SEM
the system. It should be iraructiVe to study their selection.
of files for the system and files that will not become part of
the system. Except for large dossier-type files, there should
be no particular bias except that of utility. Security should
not present any serious barriers.
The 4,mplication of the Walnut approach is quite serious.
It suggests that in principle,the notion of free index search can
be adopted throughout the community. The material in dossiers.
may then be kept at any level of classification, subject to ap-
propriate output filtering. The requester can be notified that
material is immediately available; that there are more hits, but
that he should consult the controller for permission to see these.
The importance of this techniquat)without thific7
tooksOorso access to otherwise denied files is at best a passive
matter. An analyst may send out a request for Information and
may hit a very sensitive file. Those who control the file may,
if they feel so inclined, take his request seriously, sanitize
the material, and make it available to him. This is in principle
the role that resident controllers in groups that keep sensitive
files play in substantive production areas. Given the general
SA.A_S
climate in the community, however that the
tendency to give no response in the face of great trouble mimed
be very gre1a . A detailed analysis of how such reqlr ja
handled is not available, but the Subpanel would be L1IkT to have
this point refuted.
Approved For Release 2002/05/09 : CIA-RDP80B011
pt50 050001-3
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
DRAFT
SECRET/NOFORN/CONTROLLED DISSEM
Page 54
The value of positive information about hits is that
INLA,4e"
without a need for creating a central authority 4ith the dis-
QA-0-k,
advantages inherent in ), one could insure much
greater pressure for a request to be filled. An analyst who
knows that someone else holds material hat may be interesting
/
to him, and that he has the right to request access to such
material unless positively denied by security rules, is likely
iv..
to be far more alert and diligent digging out this information
and exercising pressure via his superiors than if he is at t e, 141
1
IA',
mercy of passive responses to a broadcast req est.& Sak,"34 -kx-7-41". -11-A
kA---e-414a.Cru Ce....-.4,03,-r.AA..s -4--- citmz, kal.-0.4- (.1,,?:,t,---iy,....t. ,,,,,4a...,k
1_4)
The Walnut File approach thus offer?otential ad-
o.t.A.reuzeoi,442--
vantages both those who must maintain security of the file
and to the user who needs access to all information which he is
justified in having.
L. File Conversion
The enormous size of present iographic files create
a major problm in aut*matioLvrsjofrom menial,
partially manual, files to machine systems is hindered by the
ktaa-,3 Vo-f.
effort required
serve that the
has,
Is interesting, in this connection, to ob
pie of conversion of minimal usable files
j-iAU
4? vAR---
Cer.e.re?-
either consciously or unconsciously, been followed in some
of the experimental situations we have seen. The example here is
the Walnut/Russian biographies scheme where a name search system
has been developed experimentally for application to the totality
of a relatively small Russian biographies scheme. One of the
Approved For Release 2002/05/09 : CIA-RD
91 UW9'00500050001-3
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
DRAFT
SECRET/NOFORNiCONTROLLFD DISSEM
Page 55
problems that must be considered in doing conversion is that of
maintaining usable files at all times during the conversion
process. Th,Oubpanel noticed that some installers have
managed to convert their files or are in the prima while
others are wringing their hands and pointing to the differences
which osteneibly prevent them from starting a conversion e f t.
by
While it is/no means true that the totality of all
files should be converted, it is nevertheless true that no
fundamental technical problems prevent conversion. The Subpanel
was also L?dirig that stiff directives from high
04- .--rt--tEek ? Silf4 SQ
authority haetreat effect in
' upovercoming obs acles of all types%,-aDEQD3AeigaraF&Cpcars205ftle-
10REDMIllibp The implication Ona, 18 that the Community Is not
tibN ) c.1-k Z 4 8
For instance,
on re 13 of the CODIB Report on the
1
SC IPS Study, the following statement is quoted from the SCIPS
7
Report:
nearly as helpless maa-
"The most pressing problems of systems Integration
or interface appear to be between components within
agencies rather than between agencies."
The CODIB comment concurs, but with a hedged statement:
"This statement is probably true and deserves
careful consideration. This is not to say that
Information processing does not warrant community
consideration to a considerably greater degree
than it has had to date -- it does; but this find-
ing reflects a logical first-things-first philosophy."
Approved For Release 2002/05/09 : CIA-RDP 0
500050001-3
Approved For Release 2002/paiCIA-RDP80B01139A000500050001-3
SECRET/d0FORN/cONTROLLED DISSEM Page 56
LA-Q. Q.":70-
Ste_ 4?
clearly killing with kindness.44he biographies area
A ic ak,44-
1 is by no
more can be done within an agency than
means clear-cut that
across agency lines.
ThOSSubpanel recommends tlasilise- this question
However, IAA
p9
ODIE ability to 4; reort,(4 71
i1Wiiel adown
findings, and pull the teeth out of recommendations must be
Sk.a,
overcome during the course of Weft experiment.
tIASubpanel recommends that file conversion
to any interagency biographic file automation
cost and effort required for file conversion will
it will increase; the resulting strain on manpower
resources willigtigher. Viims a major effort should be initiated
experiments
Qv? v42-t-,---LiAtx
402
studies be int
experiment. The
not decrease
1SORIND to convert manual biographic files on
the community. The experience an
helpful in other file conversion
IP I
a large scale in
ata obtained will be very
efforts outside of biographies.
Approved For Release 2002/05/09 : CIA-RDP80W1-139A000500050001-3
1) ?
25X1A
25X1A
Approved For Release 2002/05/09 : CIA-RDP80B01139A000500050001-3
DRAFT
SECRET
SECRETINOFORN/CONTROLLED DISSEM
BIBLIOGRAPHY
"Processing of Printed Intelligence Inform:titian"
This paper was prepared for the ODDR&E during June 1961, by
Mr. while he was serving a two-week tour of mili-
tary duty with the Defense Department. (S&T Cont.No. 0553-b4)
BOREL 1963: United States Intelligence Board, Committee on
Documentation "Foreign Intelligence Objectives" CODIB-D-107/4,
16 September 1963, Paul Borel,Chairman (CODIB). Attachment:
Excerpt from DCI Report on Intelligence Objectives (Sc 08732-63/1)
(S&T Cont.No. 0229)
BOREL ;965: Memorandum for Mr. William T. Knox on "Problem Areas
of Possible Interest to Panel", 22 December 1965, and Attachment
"Selected Problem Areas in the Intelligence Community", 22 December
1965, by Paul Borel, Ass't Deputy Director for Intelligence, CIA.
(S&T Cont.No. 1861)
BOREL ;966: Memorandum for Mr. William T. Knox on CODIB Presenta-
25X1A tion to PFIAB/OST Joint Guidance and Evaluation Panel, 25 April 1966.
(SW Cont.No. 0691)
United States Intelligence Board, Committee on Docu-
mentation -rrogress Report on Foreign Intelligence Objective No.)",
CODIB-D-107/8, 9 March 1964,1 Secretary (CODIB) 25X1A
Attachment: Memorandum for Deputy to DCl/National Intelligence Pro-
grams Evaluation (6 March 1964) by Paul Borel, Chairman (CODIB)
with attachment: (Proposed Text for) "Progress Report on United
States Foreign Intelligence Objective No. 3: Automatic Data
Processing" (S&T Cont.No. 0229)
CLIFFORD: President's Foreign Intelligence Advisory Board,
Clark M. Clifford, Chairman, Memoran um for t e Pres dent on
U.S. Intelligence Community Capabi I les or the Handling of
Intelligence Information, June 15, 1965
CODIB V: Report, T/V/R-10 dated 17 January 1966 of Task Team V
(Biographies) of the Committee on Documentation of the United
States Intelligence Board (CODIB) (S&T Cont.No. 0142)(Stamped
"DRAFT") [Revise to latest version sent by Landau/Borel 6/66]
ODIB : Report T/VI/A-1, dated 28 September 1965 of Task Team
VI(Intelligence Data Handling Research and Development) of the
Committee on Documentation of the United States Intelligence Board
(CODIB) (S&T Cont. No. 1586)
CGDIB 1961: CODIB Third Annual Report 30 June 1961 Serial No.
CODIB-AR-3
CODTB-1962t_t CODIB Fourth Annual Report, 31 July 1962, Serial
No. CODIB-AR-4
Approved For Release 2002/05/09 : CIA-RDP80B01139A040Q5 000:. :1-3
;
Approved For Release 2002/05/09 : CIA-RDP80B01139A000500_050001-31
DRAFT
SECRET/NOFORN/CONTROLLED DISSEM
CODIB 1962b: Information Processing Programs in USIB Member
Agencies (Appendix C of CODIB 1962a)
CODIB 1963: CODIB Fifth Annual Report, 5 September 1963, Serial
No. CODIB-AR-5
COD/B 1964: United States Intelligence Board Committee on
Documentation "Stage 1 Report of the Staff for the Community
Information Processing Study (SCIPS)" USIB-D-39.7/5 COBID-D-82/28,
26 February 1964
CODIB 1965a: CODIB Draft Task Team Quarterly Report for the
period 1 July to 30 September 1965, Serial No. COD'S-D-112/5,
1 November 1965, Limited Distribution
CODIB 1965b: CODIB Seventh Annual Report, 3 December 1965,
Serial No. CODIB-AR-7
COD/B 1966: CODIB Draft Task Team Quarterly Report for the
period 1 October to 31 December 1965. Serial No. COD'S-D-112/6,
18 January 1966, Limited Distribution
KIWATRICK 1960: The Joint Study Group Report on Foreign Intel-
ligence Activities of the United States Government, 15 December
1960, (TS Control No. 172800) (SW TS Cont.No. 598)
=IPS 1,963: Staff for the Community Information Processing
Study (SCIPS) Stage 1 Report Vol. II, Oct. 1963 (Sc No. 12481/b3,
15 January 1964) pp. 57-70
Approved For Release 2002/05/09 : CIA-RDP8OBOV14.900t39,.
50001-3
Approved For Release 2002/05/09 : CIA-RDP80ff01139 000500050001-3
DRAFT
SECRUARZIVIMEIgLIMAt
Appendix 1
Nature of the Biographies Prob1em
Approved For Release 2002105/09 C
DP80B01139A000500050001-3
- 3
Approved For Release 2002/05/09 : CIA-RDP80601139A000500050001-3
DRAFT
S t-
3ECROJLCONThQLLED DISSEM
Appendix 2
Request for Informs, on and CODIB Support Staff ReBporz3e
Approved For Release 2002/05/09 : CIA-RDP800011 bo 500050001-3
Approved For Release 2002/05/09 : CIA-RDIVRASIe4000500050001-3
SECRET/SOFORN/CONTROLLED DISSEM 0.tbint
NATURE OF THE BIOORAPHICS PROBLEM'
"The Intelligence Community has for many years collected an
ever-increasing amount of information about individuals from e great
diversity of sources through a large number of channels, and has
stored the data in a variety of retrieval systems in diverse formats.
These have traditionally taken the form of index references, either
self-contained or leading to dossier files or individual documents.
The Team decided, as a point of departure, that the relative pay-off
in system improvement would be higher in respect to the larger
biographic files in which there is a high degree of activity and
interagency communication. Thus, many of the smaller files studied
by SCIPS (the Staff for the Community Information Processing Study)
were not included.
"There are three types of major biographic indexes and files now
In operation. They are the Positive Intelligence, Counterintelligence
and Security holdings. There is relatively little exchange of re-
quests between the PI biographic files and the Security files, moder-
ate exchange between the CI and PI communities and frequent exchange
between Security and CI. The Counterintelligence (CI) biographic
system centers around the foreign counterintelligence repository of
CIA and the domestic counterintelligence holdings of the FBI. The
Security and PI holdings of the agencies referred to in this report
also lead to CI data in some degree. The interagency exchange of
Security data centers around the name search type operations per-
formed by CIA, State, Army, Navy, Air Force, FBI, Secret Service,
Immigration and Naturalization Service (INS), and Civil Service Com-
mission (CSC). The major PI biographic records are contained in the
files of the CIA Biographic and Special Registers, DIA, NSA/Office
of Central Reference, Department of State and Air Force/Foreign
Technology Division (FTD).
"There are important and fundamental differences between, and
some similarities in, the basic operating procedures and kinds of
searches that are made in the PI systems versus the Cl/Security
systems. The PI biographic :systems are deeply intertwined with,
and in many cases actually part of, larger intelligence collection
and storage systems which are mission, subject or area oriented.
In contrast, the Cl/Security systems are clearly oriented to the
heavy use of name searching among alphabetically ordered biographic
Indexes which, in most cases, lead to dossier files. The Team
determined that there is name searching and name finding going on in
both the Positive as well as the Cl/Security activity. However, the
bulk of the requests in both areas involve name searching (above 95%
in the Cl/Security area and about 80 in the PI area)."
*(CODIB1V, p.6, Par. 1,2,3)
Approved For Release 2003/9,5nhellf-RDP801301139A000500050001-3