MEMORANDUM (Sanitized)

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP83B01027R000200110006-8
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 14
• Document Creation Date: December 16, 2016
• Document Release Date: December 2, 2004
• Sequence Number: 6
• Publication Date: April 21, 1978
• Content Type: MEMO

Approved For RdUse 2094112L22 A-RDP8t3B01027R 200110006-8
Strategic Warning St
Washington, R.C. 20301
S.-0022/SWS 21. April 1978
MEMORANDUM FOR
We believe the. recommendations in your Working Paper of
7 March address a real and intractable problem which has
never been adequately recognized or resolved. However, we
feel bound to express certain doubts regarding the assump-
tions about the nature of warning implicit in your proposals
and their feasibility and practical effectiveness. The core
of the problem, as we see it, is the perennial question of'
how to define.and delimit the task of providing advance
warning at the national level. The answers to these questions
logically dictate the organizational structures and procedures
to perform the mission.
Your recommendations treat the problem as essentially
global, open-ended, and undifferentiated.. This interpreta-
tion of the problem logically requires a mechanism designed
to deal with a formidable range of problems and events on a
world-wide scale. Your concepts of an ICWS and working groups
or task forces represent plausible mechanisms to perform a
mission defined in these terms.
The crux of our reservations is that this definition,
however logical in the abstract, is so broad and open-ended
that it constitutes not simply an expansion of the warning
function but, in fact, the creation of a new and unprecedented
mission. This mission, by virtue of its, unlimited geographic
scope and range of event coverage, would, we believe, tran-
scend the physical and intellectual capacities of mechanisms
you envisage. In view of the immense and virtually unlimited
range of problems and events this mechanism must deal with,
we have serious doubts that the end products would, or could,
meet the requirements of effective warning intelligence. in
sum, the mission and mechanism you propose would carry the
intelligence community into new and uncharted terrain and
would result in functions and products which, whatever their
limited value in identifying events which have a high poten-
tial of major concern to US foreign policy, would fall well
short of the rigorous standards of warning intelligence..
Approved For Re1ea~6--2004/12/22y:_CIA-1DP83B01027R000200110006-8
 Approved For ease 2004/12/22: CIA-RDP83BO102700200110006-8.
It seems to us that much of the discussion about warning
at the national level and the.concern expressed about the
adequacy of the arrangements created under DCID 1/5 reflect
a misunderstanding of and confusion about the intent and ex-
pectations of this directive. DCID 1/5 did not envisage the
Special Assistant to the DCI for Strategic Warning'and the
SWS as the sole-or exclusive authority for warning. On the
contrary, the directive contemplated that warning at the na-
tional level would be a shared community responsibility.
SWS was conceived as serving essentially a staff and supporting
function, and as an adjunct and supplement to the DCI, NFIB
and the NIO system in meeting their warning responsibilities.
The USIB documents whi6h defined procedures for preparing
strategic warning notices made it clear that this was to be
a broadly shared function. The guidelines stated that a
warning notice "will normally be initiated by the Special As-
sistant," but they also said warning notices may be proposed
to the Special Assistant by a USIB Principal, a National In-
telligence Officer, or by "any other senior officer of the
intelligence or foreign affairs communities." The Special
Assistant, moreover, was authorized to call on any element
of the intelligence community for support in preparing a
warning notice. SWS was assigned responsibility for coordi-
nating draft warning notices with community agencies and
appropriate NIOs. When feasible, the views of appropriate
US embassies and field commands are to be solicited.
You will note that the main thrust of our comments on
the Working Paper is an appeal that the implications of your
proposals be fully explored in the light of past experience
and of the unique demands of effective warning. The central
lesson of this experience, as we read it, is that the prob-
lems and dilemmas of warning are so formidable that this mis-
sion should be confined to a limited number of genuinely high
priority threats to US security and interests. To cast the
warning net too broadly is to risk impairing, if not defeating,
this primary function.
We believe the community made a wise and prudent decision
in 1975 to define the central task of strategic warning in
selective and discriminating terms and to invest responsibility
in the NIOs and community production offices for the remaining
broad and virtually unlimited potential threats and crises.
The total warning problem can be made manageable only through
a division of responsibility along these lines.
2
Approved For Release,20.0_411z1 22-;._CIA,RDP83B01027R000200110006-8
 Approved For RdWse 2004/12/22.: CIA-RDP83B01027 200110006-8
If community managers determine that warning on a world-
wide scale is not being handled satisfactorily, it seems to
us that the most reliable remedy is not necessarily to create
new mechanisms but to focus attention an perceived deficiencies.
in the existing structure, which already; provides for proce-
dures under NIO auspices to detect potential tension and warn-
ing situations and to organize working groups or task forces
to deal with. them. Perhaps had a sound case when
he pointed out that the remedies are "essentially not a matter
or organizational structure" but rather of warning "practices"
and doctrine.
Enclosure
Cmts on Working P.ap9r?entitled
"Warning and Crisis Operations
in the Intelligence Community,"
by 21 Apr 78
Approved For Release ,2004/11.2122 :CIA-RDP83B01027R000200110006-8
25X
25X
 Approved For Rose 2004/12/22: CIA-RDP83B01027 200110006-8 90
.SECRET
S-0022/sws
21 April 1978
Comments on the Working Paper entitled, "Warning and Crisis
Operations in- the Intelligence Community,,"' by
1. The recommendations contained in`-this paper raise
some fundamental questions about the nature and problems
of warning intelligence and organizational arrangements
for performing this mission. Together they represent a
major departure from the conclusions , of. the studies in
1974-75 which led to the issuance of DCID 1/5. The Strate-
gic Warning Staff understands the concerns that have.prompted
these recommendations for redefining the warning mission, and
we sympathize with their objectives. However, as I'and W
specialists, we feel an obligation to draw attention to some
of the implications we perceive in these proposals in order
to identify and clarify certain basic issues.
Mission
2. The most far-reaching changes proposed in the Working
Paper concern the scope and definition of mission of the pro-
jected Intelligence Community Warnin_S:taff (ICWS). Whereas
DCID 1/5 limits the mission of the Special Assistant to the
DCI for Strategic Warning, supported by the SWS, to providing
"the earliest possible warning" that the Soviet Union, the
Warsaw Pact, the PRC or North Korea "i.s. considering military
action by its armed forces beyond it& borders, or is employing
Approved For Release 2004$VZT A 2DP83B01027R000200110006-8
5X1
 Approved For Rose 2004/ / A-RDP83B01027 200110006-8
its military capabilities beyond its borders in ways that
might.threaten military confrontation with the United States,"
the Working Paper would charge the ICWS with world-wide re-
sponsibility across a broad spectrum of non-military as well
as military intentions and actions. In contrast to the preci-
sion of the warning mission defined by DCID 1/5, the Working
Paper prescribes a very general and open-ended mission: to
detect "events which have a potential impact on vital national
interests" and/or events "which have a high potential of major
concern to US foreign policy regardless of the location of
the event."
3. It is important to recognize that this definition
of the scope and nature of the mission would involve respon-
sibilities and functions that bear little resemblance to the
field of strategic warning and threat perception as recognized
and practiced over the last three decades. We do not suggest
that this fact, in itself, constitutes adequate grounds for
opposing the recommendations, but we do urge that the full
implications of this sweeping departure be carefully explored.
The studies four years ago of the evolution and performance
of the Watch Committee are worth consulting again. This ex-
perience, in our judgment, would seem to raise serious ques-
tions about the feasibility of a global and undifferentiated
definition of warning.
Approved For Release 2004c! ~1A DP83B01027R000200110006-8
 Approved For Rose 2004/12/22: CIA-RDP83B01027R##200110006-8
SECRET
4. DCID 1/5 reflected a considered community judgment
.that the intrinsic importance'and difficulties of strategic
warning were such that this mission should be confined to
those countries and developments which pose the most serious
potential threats to-US security and interests. The directive
was based on. the recognition that an attempt to deal with a
virtually unlimited number and range of problems and crises
would result in unnecessary and undesirable duplication of
effort. More importantly, a global and open-ended mission
}
would carry risks that warning would become so diffused
and distracted with areas,of secondary importance that its
primary functions could be seriously impaired. The decision
recorded in DCID 1/5, therefore, was to establish a small
interagency staff of specialists who would support the na-
tional warning mission by providing a prodding, "second-look,"
or "devil's advocate" function. The central mission of the
SWS is to concentrate on a systematic, across-the-board exam-
ination of the target countries' policies, perceptions, in-
tentions, military capabilities, and calculations of risks.
This mission was clearly based on the community's recognition
that past warning failures were not caused primarily by a
dearth of information but rather by an inadequate or incor-
rect evaluation of available information.
Approved For Release 20045,!`&fM-PDP83B01027R000200110006-8
 Approved For Rose 2004/12/22: CIA-RDP83BO1027RIP200110006-8
.SECRET
5. This judgment is shared by most observers who have
.direct experience and thorough knowledge of the dilemmas of
warning. In his analysis of the Israeli failure in October
1973, Gen. Chaim Bar-Lev concluded that "the mistake lay in
the evaluation of the intelligence data and not in the absence
of accurate and reliable information." 25X1
in his paper prepared for the DCI's meeting last month, noted
that "misperception and surprise do not usually result from
a lack of relevant-information.... In every case I have studied,
it is easy to.see in. retrospect that the relevant. information
for making a correct estimate was available..... The key problem
in threat perception is clearly the quality of the assumptions
that are brought to the information and guide the perceptions
of intelligence officers." Roberta Wohlstetter also emphasizes
the crucial importance of controlling assumptions, citing "the
very human tendency to pay attention to the signals that sup-
port current expectations about enemy behavior....Apparently
human beings have a stubborn attachment to old beliefs and an
equally stubborn resistance to new material that will upset
them .... Once a predisposition about the opponent's behavior
becomes settled, it is very hard to shake."
6. This testimony of experienced authorities is cited
to underscore the vital importance of'assuring that the crit-
ical problems of warning, presented by the behavior, inten-
tions, and calculations of the principal Communist powers,
Approved For Release 2004REA-RDP83B01027R000200110006-8
 Approved For Rose 2004/12/22 :-CIA-RDP83B01027RS200110006-8
SECRET
are thoroughly and systematically examined by specialists who
are aware of the lessons of past failures and who make a self-
conscious effort to apply these lessons to contemporary events
and problems.
ti
Functions
7. These considerations drawn from past experience would
seem to have a direct bear ing.on the functions prescribed by
the Working Paper for the "detection," component of the proposed
ICWS. This . staff. would be confined to identifying events
which pose a potential threat to US interests and foreign pol-
icy, "stimulating" interagency communications., recommending
the formation of working groups or task forces composed of
substantive specialists, and supporting these groups. The
ICWS would serve primarily as a "catalytic agent" to focus
attention on potential warning problems and to bring the
"proper talents" together. The purpose of theICWS "is not
so much analytic as it is catalytic and operational." The
actual warning assessments would be drafted by ad hoc working
groups composed of specialists on the areas or subjects in
which the potential threat occurs.
8. It is important to recognize that this prescription
for constantly shifting manning of working groups raises the
fundamental question of the kind of skills and experience ef-
fective warning requires. Is warning intelligence a distinct.
Approved For Release 2004 F -DP83B01027R000200110006-8
 Approved For RdWe 2004/12/22 :_CIA-RDP83B01027R200110006-8
SECRET
discipline requiring certain qualifications, insights, and
experience, or can this function be performed equally well
by various working groups which, given the structure of the
intelligence community, would be composed in most cases. of
current intelligence analysts?
9. The lessons of past experience again would seem to
provide some guidance.; DIA's draft report on "Preliminary
System Concept for an Upgraded DOD I&W System" (March 1978)
draws a clear distinction between warning and current intel-
ligence
responsibilities and skills. The report finds that
"A deficiency results when. the production of warning and cur-
rent intelligence is combined within the same resources. I
dications, by virtue of their subtlety in earliest appearances,
may be recognized early enough to serve as effective warning
only if they are exhaustively researched and methodically ex-
ploited. Their association with current intelligence creates
-a situation wherein only the most apparent indications may
be recognized. The: fullest implications of indications may
thus be left unexplored." The DIA draft report notes that
post-mortem examinations of intelligence failures in providing
strategic warning "illuminate endemic personnel deficiencies
in the I&W System." It points out that "although of the
highest national priority, I&W has never been formalized as
a separate intelligence discipline," and it recommends the
6
Approved For Release 2001EEATRDP83B01027R000200110006-8
 Approved For Rose 2004/12/22 :. CIA-RDP83B01027 10 200110006-8
SECRET
establishment of minimum personnel qualifications and stan-
dards for the I&W System.
our most experienced and knowledge-
able warning analyst, made some observations in her "Handbook
of Warning Intelligence" (1972-74) that are worth quoting at
some length:
Warning is not current intelligence. The best warning
analysis does not flow inevitably or even usually from the
most methodical and diligent review of current information.
The best warning analysis is the product of a detailed and
continuing review in depthof all information going back
for weeks and months which may be relevant to the current
situation. The latest information, however necessary it
may be to examine it, will often not be the most useful
or pertinent to the warning assessment....
Only in rare instances'where.events erupt very sud-
denly (e.g., the Hungarian revolt in 1956) can indications
and warning analysis be considered more or less synonymous
with current analysis. In normal times, the current ana-
lyst must cope with a large volume of paper. In times of
crisis, he may be overwhelmed, not only with lots more
paper but with greatly increased demands from his superiors
for briefings, analyses, crash estimates and the like. It
is no wonder in these circumstances that he can rarely
focus his attention on the information. which he received
last month or find the .time to reexamine a host of half-
forgotten items which might be useful to his current as-
sessment.... In addition, it may be noted that the weeks
or days immediately preceding the deliberate or 'surprise'
initiation of hostilities may be marked by fewer indica-
tions of such action than was the earlier period. Given
this circumstance, the strictly current intelligence ap-
proach to the problem can be misleading or even dangerous.
11. The concept of revolving working groups presents
another potential problem. The members of these groups, in
most cases, will already have committed themselves in their
respective agency publications or briefings to at least a
Approved For Release 200A2t22'` LEA-rRDP83B01027R000200110006-8
 Approved For Roe 2004/12/22 :. CIA-RDP83B01027 200110006-8
SECRET
preliminary evaluation of the issue or event at hand. Long
experience suggests that able'and dedicated analysts, once
committed to a line of analysis or set of assumptions, are
very reluctant to modify or abandon their views. They under-
t
standably feel a professional obligation to defend and promote
their published positions. There is also an institutional
problem.
acknowledges the theoretical value of
competing views as a safeguard against the hardening of as-
sumptions, but he warns that "one danger is that the desire
of separate bureaucracies for what economists call product
differentiation encourages dissenting estimates for the wrong
reasons." The experience of the Watch Committee in this
respect is instructive. Given the realities of group dy-
namics, it is extremely difficult to avoid resort to hedges
and evasions which dilute or obscure judgments in the interest
of compromise and unanimity, despite formal encouragement of
reports that "in ten years of
weekly Watch Committee meetings, only six dissents were ex-
pressed, even though provision for dissenting views was in-
cluded in its charter."
drawing on her long experience with
"collective" analysis in the Watch Committee, observes:.
Approved For Release 2004 TicPOfi-RDP83BO1027RO00200110006-8
 Approved For Rose 2004gjo2fDP83B01027R(f 200110006-8
It does not necessarily follow that the more people
introduced into the warning process the better the judgment
is going to be. Experience has shown that a consensus of'
all the individuals who have contributed something to the
analysis of the problem, together with their supervisors,
those responsible for making estimates, and others who may
have an interest is not more likely to be correct than the
judgments of analysts who have had experience with other
warning problems and who are on top of all the information
which is available in the current situation. Quite often
the effect of bringing more people into the judgment. pro-
cess is to dilute the judgment in the interests of com-
promise.and unanimity. Lamentable as it may be, the fact
is that the most nearly correct judgments in crisis situa-
tions over a period of years often have been reached by a
minority of individuals .... What usually happens is that a
majority in all agencies is wrong -- or. at least not right.
Thus the situation, is not taken care of by the usual device
of a dissenting agency footnote, since it will be a minor-
ity in each agency (not a majority in one) which will be
in dissent.
13. We have tried to base these comments on the Working
Paper recommendations on our assessment of the clues past ex-
perience provides for effective warning. There is nothing
sacrosanct about the arrangements established by DCID 1/5,
and we are acutely aware, as Avi Shlaim observed in his diag-
nosis of intelligence failures, that "The.search for an in-
fallible system of advance warning of an attack is the search
for a will-o'-the-wisp." However, when one examines the re-
cord of strategic surprise back to Pearl Harbor -- a record
replete with recurring errors in the evaluation of available
information -- it is hard to challenge
judgment 25X
that the required remedies "are essentially not a matter of
organizational structure" but rather of "practices" and doctrines.
Approved For Release 200 ~,q' L"TATRDP83B01027R000200110006-8
 Approved For Rose 2004/12/22 CIA-RDP83B0l027R 200110006-8
SECRET
This judgment, in our view, underscores the wisdom.of rein-
forcing the-work of the NIO system and the regular production
offices in the community by recognizing that I&W is a separ-
ate intelligence discipline and by maintaining the safeguards
represented by a smail interagency staff of trained and ex-
perienced warning specialists. The ideal "solution" of course
concept of "professional self-indoctrination"
in threat perception by all producers of finished intelligence.
But given the "real world" of the unavoidable division of labor
and specialization in the intelligence community, we believe
it is prudent to provide for a supporting "second-look" dis-
cipline dedicated to a sustained and systematic (as opposed
to an ad hoc) examination of a limited range of the most im-
portant warning problems.
14. This is not the place to propose specific revisions
in the mission defined by DCID 1/5, but a modest expansion of
the charter might be considered, perhaps on an informal basis,
to cover certain aspects of situations (e.g., the Middle East
and the Horn of Africa) which involve potential confrontations
short of direct military confrontation. It is difficult, of
course, to establish precise limits beyond those stated in
DCID 1/5 without drawing SWS into an imprudent overextension
of its primary functions. In the final analysis, the ques-
tion of scope hinges on priorities and on the community's
Approved For Release 200A9:57kv TRDP83B01027R000200110006-8
 Approved For Rose 2004/12/22 :.CIA-RDP83B01027R*200110006-8
SECRET
fundamental concepts of the nature and requirements of warn-
ing.
25X1
Approved For Release 2004$R.DP83B01027R000200110006-8