ON COUNTERING STRATEGIC DECEPTION - WILLIAM R. HARRIS

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP83M00914R002300090004-8
• Release Decision: RIFPUB
• Original Classification: K
• Document Page Count: 100
• Document Creation Date: December 20, 2016
• Sequence Number: 4
• Content Type: REPORT

? Approved For'Release 20~
UNCLASSIFIED--NOT FOR PUBLIC RELEASE
- is cocumert is being submitted to the Depart-
.iert c' Defense for review for public release.
! =ries should be directed to The Rand Corpora-.
z:c?a. Attention:' Publications Department
ARPA ORDER NO.: 189-1
3D20 Human Resourc
c Deception
A Report prepared for
DEFENSE ADVANCED RESEARCH PROJECTS AGENCY
vv CI:o m1?:' l e t e dd
7/04/17: CIAeeRDi'83M00914R002300090004-8
1 R-1 230-ARPA
Ma"Cl
SANTA MO.ICA. CA. 90+Oe
 Approved For Release 2007/04/17: CIA-RDR83FyIOO914R002300090004-8
ARPA ORDER NO.: 189-7
3D20 Human Resources
A Report prepared for
DE?FFNSE ADVANCED RESEARCH PROJECTS AGENCY
 Approved For Release 2007/04/17: CIA- DP83M00914R002300090004-8
Copyright ?.1944
Dro*ing by Chas. Addavrs;
The New Yorker Mogozine, Inc.
 Approved For Release 2007104117: CIA-RDP83M00914R002300090004-8
report was supported by ARPA Contract DACH1567CO141 with The Rand Corpora-
tion, where the author is amember of the Social Science Department.
PREFACE
When the blind seek to lead the blind, a cautionary note is in order. This report
is written in blissful ignorance of whatever systematic `efforts the United States
Government has made to construct a.counter-deceptioncapability, designed to de-
test, neutralize, and undermine those foreign deception` operations. that. degrade
governmental performance through U.S. intelligence channels. To:the extent that
such a capability exists, what follows maybe superfluous. Neither:the public litera-
ture on intelligence activities nor the :publc.record of intelligence .predictions
confirms the existence of such a counter-deception capability, but the public record
is notoriously murky and subject to correction.
This report explores the effects of deception upon strategic warning systems,
and possible protective efforts for the detection of strategic foreign deception. The
report excludes from consideration evaluation :of organizational alternatives for
counter-deception systems: Drawing only upon public sources, this report is designed
to serve as a background reference for readers interested in methodologies for coping
with deception, or in'references to the literature of deception and cognition.
Preliminary findings of this study were reported at a strategic planning confer-
ence sponsored fry the then Advanced Research Projects Agency (ARPA) of the
Department of Defense in April 1972. The report identifies methodologies ofcounter-
deceptionfplanning thatnay'be appropriate for subsequent DARPA' research. This
 Approved For Release 2007/04/17: CIA-R DP83M00914R002300090004-8.
Strategic deception has risen with the incidence of strategic: surprise in. 20th
century warfare. Expansion of systematic intelligence .collection has allowed decep-
tion planners to degrade their adversaries' predictive'peifgrmance, jeopardizing
arms control opportunities and. trust in the reliability of strategic predictions. Tech-
niques for detection of foreign stratagems are reviewed with'a: view t_o establishment
of formal counter-deception systems.
Section I offers a historical introduction to stratagem. Deception is proposed to
have been an intermittent practice in :war encouragedby (1) war-fighting strategies.
dependent upon those of adversaries, (2)=prior establishment of intelligence collec-
tion systems;.(3) asymmetries in mobility!;espionage, aerial, or electronic capabili-
ties; (4) polarity of interests; (5) personalities;'(6) deception doctrine; and (7) perma-
nent deception organizations, generally existing. since World War II.
Section II reviews evidence on the efficacy 'of deception in 20th century warfare,
primarily from the empirical historical research of Barton Whaley's Stratagem. Of
57 strategic initiatives involving deception in Whaley's study, 88 percent yield some
element(s) of surprise; amphibious landings associated with deception yielded sur-
prise in 89 percent~of cases. Although it is widely assumed that intelligence predic-
tions have been hiipros ihW, given radar, better overhead reconnaissance, and other
techniques developed in World War II, Whaley found that fully 68 percent of land
battles in World War n yielded surprise, compared with only 48 percent in World
War I. The Whaley datum inay underestimate both the incidence and the usefulness
incidence & siiiprise, Whaley tested the effects of the intensity of deception upon
of deception in war. Data on high-intensity deception operations were segregated,
from other, cases; in 27 of 28 cases some element of surprise was obtained.
Following this report's testing of the effect of high-intensity deception upon the
:``~aualty ratios'aad found that high-intensity deception correlated with reduced
ations resulted in some mode of surprise, 96 percent or more of high-intensity
casualty ratios- of cieceiver to deceived for Lnose aLLacxuig wii.n uruau ? cu Ver uuu
deception;: Whaley found that increased intensity of deception correlated with in
creasin` intensity of surprise, but not with an increasing overall incidence of sur-
prise. This report's findings indicate that increases in intensity of deception corre-
late with both increased likelihood of high intensity surprise and increased inci-
dence of surprise in one or more modes: intent to attack, time of attack, place,
strength, or style. Whereas 80 percent of low-intensity strategic deception oper-
operations resulted in some de ree f su17risew._?_
 ? i- ipi uv u rul rvCICdJC LUU! tU'-ii I I . l~lt1-fCUr'OOIVIUUZ7 i'-ir uULJUUU~UUU~F-o
17. ~It is onc~lud tnat in the assence of=counter eception systems, high-intensity -'
deception has resulted in almost certain surprise, with great damage to defenders.
Although the Bayesian probabilities from nonnuclear warfare should not be directly
applied to estimation of prelaunch warning in nuclear war, data on past incidences.
of surprise do not generate a priori confidence in the reliability of prelaunch strate-
gic war warning systems. -'.'.
Asymmetries in intelligence capabilities, particularly in counter-espionage
duels for control over adversary collection systems, have contributed _:not. only to
F high incidences of surprisebut also to surprise in the costliest cases and modes:Also
contributing to the failure of intelligence predictions has been the bimodal informa-
tion theory conception of "signals" and "noise," which leaves no room for the spuri-
ous signals, or sprignals, designed to penetrate the filtration barriers of signal`
processors. ` - . .
The key impediment to strategic arms control verification.is the challenge of
deception. Section III addresses methodologies for , couunter-deception planning. Ad
bocefforts of intelligence analysts to cope with deception . have. been notably -unsuc-
cessful; judgments of captured enemy war plans have been significantly worse than
d
as
would have resulted from coin tossings. Four of five genuine plans were treate
uine..:~ .
forgeries, five of five deception plans were accepted as gem
Three techniques for the detection -of deception are outlined the reconstructive
inference method, in which the reconstruction of sprignals patterns is attempted;
the lncongrwty testwgmethod, the tad tional form of intelligence analysis, which
tests alternative hypotheses; and the *: vulnerability assessment method, which in-
volves the prediction of future vulnerabilities on the-basis of Bayesian estimates of
conditional probabilities derived from past, cases of deception.
National styles of deception are germane. to both the reconstructive inference
method and the vulnerability assessment approach. Previous Soviet "war scare"
practices, which spoof strategic war warning systems, are inimical to arms control
and war avoidance objectives. Traditional Chinese stratagem has aimed at winning
without fighting, : by.manipulating the utility functions of the adversary so that he
? confounds himself.. Communist stratagems in the Chinese civil war relied on strate-
gic movement, with lures, traps, and multiple feints. The preponderance of Soviet
and American strategic-power, combined with a sense of territorial insecurity, may
encourage Chinese strategic force procurements that emphasize mobility and strata-
gem-and.could create innumerable intelligence puzzles. The understanding of past
national -deception practices may contribute to Bayesian formulations of attack
based anon a priori practices and probabilities.
-,traditionally. reliable sources or compartmented intelligence as test standards.
Channels: that"are unquestioningly accepted are favorites of deception planners.
(Probabilistic information processing, cross-testing, and retesting alternative hypo-
'tlieses'are:preferable to testing against a subset of supposedly reliable data. Decep-
tion planners have often.14 nforced'. preconceptions or anticipated expectations.
Prominently displayed self-deceptions maybe reinforced` by foreign deception plan- -
ners.
Counter-deception: planning involve's not'only detection of deception but also
countermeasures for stratagems that have been detected and those stratagems that
d
d
.
etecte
may be anticipated on the basis of past experience though not specifically
 Approved For Release 2007104117: CIA-RDP83M00914R002300090004-8
T Detection ofstratagems is complicated by randomly generated feints, just as strata-
gem is complicated by randomly generated countermeasures. In view of high inci-
dencrs of-deceptively induced surprise, designing strategic force postures that are
not substantially degraded by surprise attack is a central element of counter-decep-
tion planning.
Among counter-deception countermeasures are: (1) modification of iiitelligence -
collection systems to provide secure, variable, corroborative data; (2) pre-crisis re-
source allocation (some of which can be determined by operations- research-and
-- - gaming) to minimize the costs of being deceived;.and (3) the mounting of counter-
stratagems (stratagems that play upon vulnerabilities in adversary stratagems) -:-
once deception has been detected.
Section N addresses problems"of deception and counter-deception planning in
an arms control environment. Deceptively induced uncertainty may contribute to
mutual deterrence, particularly as surveillance and guidance systems become more
precise. Some decoys may be more valuable than: actual weapons at lower cost,
because the resulting uncertainties complicate attack planning. Deception may
prolong the survivability of existing strategic weapon systems..Counter-deception.
systems should be sufficiently discriminating that they can identifythe approximate
purposes of detected stratagems, thus avoiding unwarranted pressures to abrogate
agreements merely because enemy "deception" has been identified.
Because the record, of 20th centuiy';wartime:'deception indicates considerable
advantage to attackers, counter-deception-systems may reduce both the likelihood
and the intensity of advantage resulting f-om deceptive" attacks Insofar as military
attacks are surprise-dependent--to be launched only if surprise can be anticipated--
enhancement of strategic warning is a contribution to deterrence. If mutual fear,
distrust, and recrimination are detrimental to a detente in international relations,
of which arms control is Jost one aspect, enhanced confidence in the predictions.of
the world's leading intelligence services should provide statesmen a firmer base for
international commitments..:
Special efforts . to detect adversary' deception activities may also contribute to
arms control measures;that have been blocked by past vulnerabilities to deception.
On-site inspections whose patterns are known to host states and "facilitated" nation-
al verification in known patterns are subject to deception, termination, or both in
that order. Counter-deception- systems may turn adversary deception operations
into early warning systems, providing safeguards in arms control arrangements
beyond the value of on-site inspections.
? = : in the .late .20th century the communications systems of all the major states
could be manipulated by foreign stratagematists. Enhanced counter-deception
-.-capabilities may restore a measure of integrity to transnational communications.-
.Maa::is not.alone in the use of deception, but most studies of other species, social
insects .and animals, suggest use of deception age-inst other species. Evolution of
closely linked genetic traits, permitting the mimicking of nonedible Species'- traits,
suggests:a continuing role for defensive stratagem in natural: selection.
In 'certain' species primitive cou ter-deception measures against predators have
been reported. It is not yet known whether other species.avelearnedthe control
of deceptive capacities within the species. If they have, perhaps man, too, may learn
to cope with his apparent skill in deceit. And if man's intellectual powers have
permitted previotsly unknown deceptions within the species, those same powers
should permit establishment of counter-deception systems that may render such
stratagems less likely and less deadly.
-RDP83M00914R002300090004-8 -
 Approved For Release 2007/04/17 CIA-RDP83M00914R002300090004-8
I wish to thank Thomas A. Brown, Merton E._ Davies;'"Anna S. Ford, Timothy
Hallman, Francis Hoeber, Arnold L. Horelick, Fred C.'Ikle,;Brian Jenkins, Amrom
H. Katz, Robert M. Komer, Joseph E. Loftus, Andrew :W. Marshall, Lucien W
Pye
.
,
and William W. Whitson, who shared time and thoughts on topics of interest to me
in the preparation of this report.
Professor Barton Whaley of the Fletcher School of Law-, anti Diplomacy has
exchanged thoughts and references on.thiss?`subject. He has been kind enough to
review my critique of his provocative study and to continue a fruitful dialogue on
this subject.
Professor H. Wentworth Eldredge of.the*Depaif .ent.fofSociology, Dartmouth
College, offered valuable observations aiid criticisms of a preliminary draft; Dr.
Richard Warren of the Instrumentation Laboratory, M.I.T., offered comparisons
with ornithological and mammalian species'.. deception patterns; and Dr. Ladislav
Bittman provided commentaries on his previous work as Deputy Chief of the Disin-
formation Department, Czech Secret Intelligence Service (STB). I am indebted to all
of them, and to the'staff of Raid's Social Science Department, including in particu-
lar editorial assistance from Christine D'Arc, and secretarial assistance from Marjo-
rie C. Schubert:'
The New Yorkerhas allowed reproduction of Charles Addams"'swinging" car-
toon from its July 1, 1944 edition.
 ACKNOWLEDGMENTS........... . .......:........... ix
1. A HISTORICAL INTRODUCTION WTO:THE ART Or-
STRATAGEM .........
Notes..... ............
......................................... 5
Approved For Release 2007/04/17: CIA-RDP83M00914R002300090004-8
CONTENTS
FRONTISPIECE . .............................
PREFACE ....................................... ... V
SUMMARY ..... ....... ........................
...... ......... Vii
The Nature of a Counter-Deception System ........................ 33
Seatof-the-Pants Counter-Deception Methods ...................... 34
Detection of Deception: Three Theories ............................. 34
Reconstructive Inference ........................................... 36
:.Testing :for Signals and Sprignals ................................... 40
'Incongruity Testing.......... ................................. 46
Guarding Against Expectation, Self-Deception, and
Commitment .................................................. 49
Vulnerability Assessment ........................................ 50
The Style of Soviet Stratagem .................................... 56
Chinese Styles of Stratagem ...................................... 56
Countermeasures ................................................ 60
Notes ........................................................... 63
II. THE EFFICACY OF DECEPTION...." ...........
The Aura of. Secrecy .. ....................................... 10
A Reappraisal of the Whaley Data ................................ 12
Models of Erroneous Intelligence:... .............................. 21
Speculations. on the Vulnerability of Arms Control
Agreements.--..-.'-.--,-, ............................................ 22
Notes........::::.:::...... .24
.................................. ..
III. COUNTER-DECEPTION PLANNING: METHODOLOGIES ......... 33
 Approved For Release 2007/04/17: CIA-R.DP83M00914R00230009000478.:
IV. COUNTER DECEPTION PLANNING IN AN ARMS CONTROL .
ENVIRONMENT ..........: ................................. ,79
Deception in the Service of Arms Control.......................... . 79
The Necessity of Counter-Deception Capabilities ... 80
Stratagematic Vulnerabilities in Arms Control Environments ... .:~ 82
Safeguarding Verification Systems ..........................
The Control of Deception: Is it Possible? ................... 83
Is Homo Sapiens a Deceptive Species? ....................:: 84
Notes................................................
a:':3?
 inferences.'
Approved For Release 2007/04/17: CIA-RDP83M00914R002300090004-8
At the outset, strategic deception should be distuzguished from propaganda and
disinformation, to delimit and emphasize the source ?of present concern. Strategic
deception is the intentional creation of false reliance within-:a'-'.target group, to
mislead that group for strategic purposes. This study will corisder:strategic decep-
tion of foreign targets, generally coordinated by intelligence or security services and
directed at foreign policy elites. Propaganda and?associated psychological operations
are directed at wider targets, often by:, ore diffuse :channels than those used in
strategic deception. Propaganda, both off cially acl nowledged and covertly present
ed, is used in deception, along with false.`infciraiation (disinformation), physical
camouflage,.display of decoys, mounting of physical feints, transmission of"dummy"
radio signals, and the leaking of true informatioiipackaged so as to encourage false
Deception does more ~than'replace passive "security," a shield rendered inade-
quate by the magnitude and penetration of modern intelligence. Strategic deception
in its more ruthless -aspects"yields:more than uncertainty and the consequent.
spreading of enemy resources, skillful deception causes a redistribution of the adver-
sary's resources mi 6e' wrong directions, thereby assuring not only surprise but its
full exploitation.
A recent study, Professor :Barton Whaley's Strat gem:,Deception and Surprise
in War(1969), has introduced an expanding literature on strategic deception in 20th
is strategic`deception of unrelenting importance in all warfare, or are the tech-
. U { LLC
importance ofinethodologies to counter its effects, I shall discuss a few practices of
-:deception'-in-military history. -
ques of stratagem favored in some eras more than in others? On the basis of all
? :~.too fragmentary data, I shall attempt to identify some of the conditions under which
deception;; operations have flourished.
Tere:has been intermittent, not continuous, practice of deception in war.
This, not a startling statement, but it is a necessary prelude to analysis of the
conditions that make stratagem opportune. There is a tendency to legitimate a
practice of the present through discovery of a historical tradition. But to write that
"deception runs through the history of war,"-' or that there is "ample evidence of
[deception's) use.throughout history,"' is to discourage analysis of the discontinui-
ties, which are more striking than the continuities. Professor Whaley notes that
deception is "infrequently, or rather, intermittently and idiosyncratically prac-
STRATAGEM
 Approved For Release 2007/04/17 yq-RDP83MO09'14R002300090004-8
ance from the war-fighting practices of Bronze Age Greece?}-There is little evidence
of deception operations.in the warfare of Medieval Europeanother period of defen-
sive fortifications. And the codes of chivalry that accompanied feudal warfare were
inconsistent with th:; systematic practice of deception: _On?~the Western Front in the
first years of World War I, the regression to the phalanx in-:trenches, to attrition
warfare along fixed fronts, did not encourage the practice'.ofdeception.
emphasized the solid wall, or phalaax.Two walls of men confronted each other, often
the more numerous group prevailed because its phalanx outstretched that of the
opposition, outflanking it. There was little need of intelligence and little: use- for
ti=
deception .6
During periods of defensive fortifications and "siege warfare; victory', was.
gained through logistical preparedness in defense, scaling or bombardment. tech ~.- -
nology, or endurance in attack. The description of the deceptive attack from the -:.; .:-
Trojan Horse probably survived several centuries of oral legend because of its vari: .-
tired," and that "deception is one of those odd strategic techniques of war.. tliat.
seems.fated to cycles of loss and reinvention.""
The history of stratagem's neglect is not a topic of rigorous investigation. Archa-
ic warfare, in Assyria, Egypt, and Greece--as depicted on monuments and vases--
enemies or walled city-states prepared forsiege, fhe incentive to initiate stratagem
is reduced.
The cyclical pracecs by which deception is practiced, 76W tbie4 rediscovered is
pot a random prod environmental factors encourage deceptioq practices.
Among these are the following:
First, interactive strategies must :pre-exist. Tlie incentive to deceive an adver-
sary is quite obviously related to the iralue of any, exulting misdirection. Where
action is substantially independent, as wiith: nomadic tribes who fled from potential
Second, the prior establishment of intelligence collection systems has en-
couraged deception. By an .".intelligence system" I do not mean the mere presence
of occasional spies,. Rather ?I assume a climate of rationality and planning tt::tt.
nurtured systems-of : planned intelligence collection. Such systems have been is
evidence, with . occasional;lapses,'since` the Renaissance, most notably in 15th cen-
tury Venice, in postal intelligence systems of the leading sovereigns, and in the 16th.
century organizat obr 'af.the.':British. secret service. Although the history of pre-.
i
Renaissance intelligence -has'not been the subject of rigorous study, scattered inves-
tigations suggestthe following. unanticipated hypothesis: Deception seems to corre-
deception operations than with intelligence efforts of the target state. The mounting
of deception. operations may involve a substantial element ofassurance that one can
late. better. with establishment of planned intelligence systems by the initiator of
redid favorable: deception targets and outcomes.
deception: andthe existence of intelligence systems sponsored by states that are the
targets of adversary intelligence. For example, Byzantine commanders of the 7th,
sth_'.arid:9tli centuries A.D. probably launched the most active intelligence and
=` '.:There appears to be a weak link between the reported practice of wartime
deception *operations;of.t e., periods. but their main adversary and threat to the
continued existence of the Empire, the Arab armies,"also possessed advanced intelli-
gence systems. The. Arabs,,by this period, were working on the decryption of foreign
ciphers,' a practice not known to occur in otliersocieties "of this period.'O
On the other hand, there are examples of commanders wlio appreciated both
-,,intelligence and deception and managed to deceive even those who were not listen
 Approved For Release 2007/04/17: CIA-RDP83MO09-14R002300090004-8
.3.
espionage, aside from its appreciation of deception, in the 5th or 4th`century B C.1
The Roman texts of Frontiaus,. about 90 A.D., and Polyaenus, from : the Parthia
War in 162-165 A.D., follow development of elaborate scout intelligence systems in=
the Roman empire, as do the fragments of Polybius' Histories:relating 2nd century
B.C. Roman encounters with stratagems of the Macedonians .The Stratagemata of
the Byzantium Emperor Leo (either Leo VI, the Wise;'-.:or.Leo II, the Isaurian)
corresponds to the best ongoing intelligence establishment-of the'1 edieval period;
and the resurgence of interest in the classical deception:lituratuire, dating to the
translation of Polyaenus into Latin as Strategica in1549 A Di:follows the resur-
ing. Hannibal, during his invasion of Italy in the 3rd century .C., plundered the
countryside, spreading news to his Roman counterpart and drawing Roman forces
behind him into an ambush at Lake Trasimenus. The Roman forces failed to send
advance scouts but managed nonetheless to be lured into a trap.
In general, intelligence systems flouished before and during periods of great
interest in stratagem. Each ofthe extant classical texts on stratagem arises in a:tiiae
and place in which intelligence systems were strong. Sun Tsu's Art of Warfro n the
Chou dynasty in China indicates a mature perspective on intelligence and. counter-
.. 't
codebreaking observed in Venice from about 1400 onward. ` ` ;.=
The availability ofasymmetries, or targets ofopportuait? y, have encouraged the
practice of deception.
Beyond the necessity of a strategy interacting with that of one's adversary there
are the temptations to press opportunities of asymt. E E E'?a' E L N V >. >.t
V (??. G i-? L? ~.? , C] '?' ~. C L Y L E L V O L CO CO V
U+ a+ U \.\ \' >`\" CA a >. V V V I. V U C C\ C 0
W m
E \ >. :-% >..-I. >. 7 E .? s U U U V. U\ a s >. a ]G \ a \ N
N
L L C. >. C C. C a G\ a\\\ U\ . ea ac o. -4.0 ??4 V
L a V O.' V m. a s 'Y" q i.C V Y Y x x\ X L a a a a V= d\
itUUJ~C 'E E r--8-a 0 ?+}a7 rGOt+r '7 "i E'7Z aU acG
+\\\?Y?' L \ Y. C6 + + + N + + \"'s L\\ L\ L CO
Yy N"y, V ~~,1r V .a N N N N N N N N C. X N aC N 0 N
Y1.C:~ l0 tD ?e.1'O. Q?tV ?O''.i ?.d? H C?1 .7 .7 .T .7.a . .T .n N n C 4 .O ?4 i CO
.C .4 ..4 4"C .14-7V .T ~T Q:v .7 .7 v V V .T .T.T.T .7 7m l.O .o .O
CT O. O.?O. U C, "O. I O. 0% C. O. -M 0% P o. O. O. O. CT C> O. C. C. Q. CT O. 0.
.4 .4 .4 ..C..i .e.:?1 O ?4 .4 .4 .4 .4 .4 .4 .4 .4 N .4 .4 .4 .4 .4 -I .4 .4 .4 .4
S. Y O. CL CO.6 ?C.. a..'0 C V C 9. c C ?4 CO Y C >..3 0. Y L C to
0. U a Cl 7;'V..0..4 0:7 V V O a 3 D 7 D D V a r V a u 0. 7 7
- rx -. ni :;
?.'.{: Y .~ C Y to s >. 0 >. V 0 C-o a C. C
0..J- C O .C -C W CC S E .1 0 0 C M a v a U >v ..
0.'t$ 0 0. CEY. 0 W CO O >. Y C 7 C C IS M a O ???. a
.4 4 .4 -4 0 .~?C 7 . C.0 a .4 C > 'a V a < u N .8 &. .4 U. >. N > Y i/ C >. N V
:.w Y Y V 0 a c ?4 7 V .4 .C u O V 0 . a O a b. C .4 a -4 N
;C3m 0.OMNR't~. 47 Us 17Z Co .T. N.i IQ HM to NC,l
.... :.4.0'/. ri V1 CO .O Vf CO 14 .T Vl".O ?C!
. co 0% 1" ~L=O- n N .O CS
!V .O CO d.?-4 -4'.4 84'4 Cv 141 P1'C-1 J ?P .T .0 -7 .T V...r~.
.C.< < < < < C 6 Y C
Y aaa u W
c I c 0
o C V .4
eosin d
00814 Y
a a?0 Cl
?nsaCa .~
Y 6 .C Y
W .4 L 0
O a a 0. a
u .: Y C
V C O 0
u .a s a .4
C 0 Y - B
a a V
O CA .C C)
+ Cl C. CO
0 0 C# 10
CC... to
?
0. Y
0 Y V V O ~+
a .8 CO CI
T W Y . .0 y
a W Y U
96
a yi. O C. A
.7c d n C 0 O.
Y 74 O .4
Y 1. 0 p v
Y c u%
p v m . .
0
Cp a a a E +O+
5-44> u a L
a .r a m C H
CCC?.
M V .O s
a)s >..a .C?+ a
.4008 Li ->
~:+o o me
m .c u a Y_
.4 0 7a M is a
W C a w ??
.4 a .- .4 . a
r 0 .0 LA
..4 v v a 4 >
u! V pcp > x O 0 it 0 8 96 x a
 Approved For Release 2007/04/17: CIA-R DP83M00914R002300090004-8
17
Assuming the data presented in Table 1 to be reasonably accurate, we are`li
Also o `-iateresti are the following statistics: First, the average magnitude of
surprise in these 28 cases of high-intensity deception (using the Whaley index,
adding the various modes of surprise in each case) is 3.0, well above the average for
Whaley's remaining 40 cases, 2.37; second, 43 percent of these 28 cases (as opposed
to 37 percent of the other 40 cases) involved the misjudgment of basic intentions
(attack vs. no-attack), probably the most consequential and costly aspect-of surprise;
third, casualty ratios of target state to deceiving state in these "grand-"" deception
cases (median = 4.8 to 1) were far more favorable to the deception.sponsoringstate
== `='
than in the average case of deception or surprise attack.
resources.
which there is little evidence-that the originator of a strategic initiative willing to
expend substantial deception resources can expect nearlya 100 percent probability
of achieving strategic surprise. Further, he can desigrra-set. ofmisleading indications
so as to capitalize upon enemy surprise, by encouraging misallocation of enemy.
perations areidden from public view,
If some of the more diabolical deception operations.
the efficacy of deception may yet be understated. In critical gases, the estimated
likelihood of surprise (0.96---) could notaise much higher, but=new details of the
misallocation that resulted might make -_i U- prosPects of deception appear even
more grim.
Reviewing 168 battles in 16 wars j~iii #lie 2dth century= Whaley found that of 59
battles fought without initial surprise: only.t~vb:percentlexceeded the battle-initiat-
ing leaders'-expectations, and 60 percent of'these battles resulted in abject failure.
'
objec-
In contrast, of 50 battles involving surprise -percent exceeded commanders
tives and only two percent ended in failure'g-These subjective evaluations are
consistent with empirical; data-.'on battle casualty ratios. Whaley studied casua',y
ratios in battles undertaken with or without surprise (eliminating the upper and
lower five percent .of'casualty.*.ra tio. cases). Without surprise, in 45 battles, mean
casualty ratios .(attacker/defender). were 1 to 1.1; with surprise, in 79 battles, mean
casualty ratios were`1. to 5.3, somewhat higher in tactical encounters (1 to 6.2),
somewhat lower in major battles (1 to 4.5)."
A a might be expected,.increasing intensities of surprise (counting five modes: go,
no-go; time; place; strength;.end style) produced more favorable casualty ratios:1 to
-.11.5- i all five modes ofsurprise present (still not counting the five percent ofcases
1 w. inf World War L Of 61 cases of strategic surprise, 82 percent were found to involve
t chances of attaining battlefield objectives, and a 34 percent chance of achieving
objectives well beyond those antici ated.
=with Iargest-sualty ratios), higher absolute numbers of casualties by the surprised
`; forces, lower absolute numbers of casualties and higher numbers of prisoners taken
?
by the forces- initiating surprise."
. .. _ w w ?~.t" .1 a it _?, _ _ _ _ _!?___ __ ,,,.. __._L._~_ -~.S ~~ t.....
involve 'strategic surprise. Crude measures of high intensity deception indicated
that the concentration of deception resources in key battles correlated with even
-ategic:.deception, and 88 percent of cases of strategic deception. were found to-.
higher incidences of surprise, 96+ percent. We have also seen that the infliction of
t high intensity surprise resulted in more favorable casualty ratios, reduced casual-
ties for the surprise-initiating forces, more captured prisoners, much improved
 Approved For Release 2007/04/17: CIA-RDP83M00914R002300090004-8
No decept;on
cases (n = 73ruse index-6)
Fig. 2-Effects of intensity of deception upon surprise
constitutes surprise I wish to treat with caution hypotheses lucking the intensity of
deception to the intensity.=of surprise. Figure2 adapts data on the relationship
between. intensities of~deception and surprise as recently reported by Whaley"
Initiation of at leaka.low-intensity deception operation correlates with substantial-
ly higher incidences if suipriseJn, 20th century warfare than would result from
attacks withoutassociatedddeception efforts. Moreover, it appears that an increased
deception effort a ensured bythe..aumber of deception modes used,correlates with
an increasing incidence. of, high-intensity surprise but not significantly increased
prospects of achieving.-surprise. The shift is from low-intensity surprise to high-
Given the imprecision of evaluatioris;of what constitutes deception and what
Whaley's data, displayed in Figure 2, are not entirely consistent with my own
data?-'oii tlie:relationship between the tntensityof deception and probabilityof sur-
prise .derived from the same Whaley case studies. Whaley's findings suggest, in
intensity surprise, whi&?is'more costly to adversaries.
:-ffect, that switching from low- to high-intensity deception operations correlates
ty of surprise.
?.Witn increases in the intensity oz surprise Out no signincant increase in the prooaou-
,'=As:dist~is,rl earlier, my review of Whaley's case studies of surprise in strategic
battles-:(mainly commencement of war, opening of new fronts) suggests that switch-
operation"s'(riise`indea of 4-8) resulted in some degree of surprise, and all strategic
deception cases with a ruse index ?- 5 resulted in some degree of surprise.
inteiasity'aiid the "probability of surprise. Thtts only 80 percent of 30 low-intensity
strategic-deception operations resulted in some degree of surprise (ruse index. of.
surprise `index of 1'-5), 96+ percent of high-intensity strategic deception
mg'frcitn?`-l w 'to'high-lht;6hsity?deceptioni correlates`tvitl increases in both the
 Approved For Release 2007/04/17: CIA-RDP83MOU914R002300090004-8 -
19
. My findings suggest that, in the absence of counter-deception systems,-high-
intensity deception in strategic operation's should not only result in a higher-inci-
dence of high-ittensity surprise but also is a reduced prospect.of failing to achieve
a significant element of surprise. If it is easier to obtain an element of surprise in
tactical battles through low-intensity deception operations than is possible ie strate-
gic encounters (of more central concern to adversary analysts), then my firidiags and
those of Whaley may be consistent.
Diverging conclusions might be Important in estimating whether a nuclear
attack were possible without a substantial risk of pre-launch disclosure 'of the. ;A4
'
tacker
s intentions. If one concludes that high intensity deception can drive the:
probability of pre-launch surprise close to 1.0, rather than 0.8 to 0.9, one may reach.
different conclusions about defense postures required to assure nuclear deterrence.
Without further research, however, we cannot assert that. high-intensity deception.
operations can drive the probability of achieving surprise close. to 1.0; and in any
event, we are wary of applying the Bayesian data on conventional warfare to the
probability of pre-launch warning f nuclear attack:;=Nonetkieless, the empirical
findings on the relationship between -deception and 'surprise;. and between high
intensity deception and high-intensity surprise, do not generate'a priori confidence
in the reliability of strategic war warning systems.
Relying upon unreinforced adversary preconceptions, operational security, or
luck has yielded an incidence of surprise of only 27 pe erentlyrcenta0 Stated diff , the
chances of failing to achieve any significant elementof surprise without deception
:ofinternatiorialcomduct; it humbles all and affects those who are not its targets but
"Iai _e nations and'smaiiandbetween rich and poor. Deception is one of the equalizers
public literature on intelligence forecasting hass failed to address the challenges of
counter-deception planning so as to provide more reliable strategic warning systems,
and in the process to discourage would-be attackers.
What of asymmetries ia:strategic deception practices? In. Section I, I suggested
the same states that-foster rational* planning of intelligence collection may tend to
appreciate potentials for deceiving adversaries through their intelligence systems.
Those states that .have succeeded .inpenetrating the security of adversary systems,
placing agents well coding high-level communications, or controlling adversary
intelligence systems have also gained insights into the decision styles of their adver-
saries, ,the concerns, prejudices, assumptions, inference processes, channels, and
substance of information tliat are ripe for misdirection.
the preceding statistics tell us that even the disadvantaged have opportunities
for . surprising- their adversaries. The incidence of surprise has been high between
are almost 3 to 1. That deception has become part of the'niodus operandi of modern
military planners is entirely understandable It is less understandable that the vast
who: fear its'consequences..Available statistics do not convey the effects of asymmet-
esources We may suspect,. but cannot as yet prove, that those states with
not allow empirical study ofthii consequences of intelligence asymmetries.
.costly'to'an aiiversary Without mbre`de'ta led-- public evidence, the data base does
deception allow greater opportunity for the kinds of misdirection that are especially
irifensity surprise The opportunities for either massive deception or sharpshooting :
'convert more--ca ses of what would be low-intensity surprise into cases of high=
adv-an-tagepf intelligence or control of adversary intelligence systems may manage
Asymmetries have existed and have tempted exploitation. Sir John Masterman
explains of the British-German counter-espionage duel of World War IL
 Approved For Release-2007/04/17: CIA-RDP83M00914R002300090004-8
vea
s
Y: a DUO-cep ty ofcommaaders to_pasticular packages
of"signals" and to the channels by which they are. transmitted :When the deception
planner has at his disposal well placed networks of-doible agents, he has ready
access to the collection system of his target. If he is careful; he wl,cover the other
- r- ---~- -?~..,~ ~.~a~. au agcy w ule cues anu;cipners ox tine opposi,-.
tion's intelligence services; second, access to the command?'decisions as transmitted.
The first reveals the collection patterns, inferences, preconceptions, and fears of the
enem the second
l
th tibili
patiently cultivated network of double agents, deception is almost?a'sure Winn r,:
If we assume a signicant disparity between the deceiving state's intelligence
resources and those of the deception target, with the imbalance favorable! #o: the
deceiving state, we can probe strata$eraic vulnerabilities in their rzx i ,irulent
form. Deception operations mounted without the assistance of "magW--,. cry tologic
intelligence) or well-placed agents are a gamble. When supported .bymagic:snd'a.
Not only have double.ageats been run on a long-time basis but they have
been run, so extensively that we can think not in terms of a number of
isolated cases but in terms of a double-agent system.... We actively ran and
controlled tie German espionage system in this cowztry. 1
This unbalanced situation is the stratagematist's equivalent'of"pay 7V." The
adversary pays for the opportunity to:* " ai y bf several images, transmitted to him
over various channels designed by the stratagerii4i! t'Withaccess to the preconcep-
sensing systems as well
mental deception cornmittee':discussed the "fine tuning" of deception programs in
the light of intercepted Abrvehrintelligence? Certainly we can. And we should not
be surprised. to learn in Whaley's Stratagem that Abwehrintelligence was not only
atiorig . :most of the critical cases but also wrong in the costliest of directions.
unlit tie perceives the incongruities he willbe.satisfied.) Where the stratagematist
has access to the internal channels of his adveisary's intelligence, he is able to learn
when his target audience is:not-interpreting'the various images as they were meant
to be read. If he can adapt his programs in mid-course, the stratagematist can almost
guarantee that the , audience is both satisfied and convinced. In short, with "magic"
and double agents: the stratagtmatist can even control the '.`fine tuning" of his
audience's perceptual: Hugh Trevor-Roper; tells us that the British decrypted Abwehrciphers almost
straight through World. Wa _11.42 Can we believe that at 10.00 a.m. an interdepart
:':has not exisfed:during critical wartime threats or else it prefers to skip over "quirks"
':when totaling the batting averages.'' The fact that most intelligence services are
%
bons, values, and channel preferences of his viewer,: the=stratagematist designs a
seri es of
pmgrartss that willsatisfy the ciistoiaier iThe customer may be misled, but
badly deceived at just the time when they are most needed is, not a.valid argument
or their:demolition. The smaller,,less"sensational successes may, entirely justify the
investzaens;.further, development of, some. counter-deception capabilities may at
least ameliorate. the rather poor track record in the most important cases. Deception
security assures that the track record, though obviously bid, is ascribed to fortuitous
circumstances, random developments,.poor,luck, the inherent. unpredictability of
most situations, and a host of other circumstances:
 Approved For Release 2007/04/17: CIA-R DP83M009.14R0023000900.04-8
MODELS OF ERRONEOUS 1N'TEU IGENCE
:'adopts thus mode of analysis, suggested in 1967-1968.?8 So do his less rigorous but
Omitting the visibility of deception planners, the predictive records of intelli-
gence services still look bad (for the evidence of resource misallocation is. often
blatant) but their rational exegesis seems nigh impossible. The evidence is .aiabigu-
ous, the warnings hidden or late in arriving. (How many a deception operaovr can
remember passing a correct warning just before the battle, too late for'itaexploita-
tion by the enemy but sufficiently early to accredit the sending channell-for?later
play?)
The most persuasive model of this discouraging state of affairs is finis that;._
accepts this abysmal record as the natural if disordered circumstance of intelligence-
work. With hindsight, the inferentially valuable data..are.identified as a group of
signals, the troublesome and misleading are identified as.noise. With such a model,
post mortems of almost any of the major turning points i -20th century warfare lead
to the conclusion that separation of crucial signals: fthrn the ambient background'
noise,is extraordinarily difficult, that potential improvement is highly restricted,
and that the possibility of calamitous surprise is everpresent.
The outstanding case study of this genre is Roberta W:ohlstetter's Pearl Harbor.
Warning and Decision 44
A study that followed Mrs, Wohlstetter in merging the
deception data with background noise.-is .Holst's analysis of the-German surprise
attack on ' Norway in April 1940.*' In th:e case: study. on - Operation BARBAROSSA
that led to his monograph on Stratagem, Whaley:incl ided the signals in a draft
chapter on noise until he realized tl at':most 'of-'tile :noise was manufactured in
Germany, with specifications designed .to -penetrate Soviet intelligence filtration
barriers'e David Walker's Lunch With a Stranger notes that it was the British.
practice in World War II to call the output;of :controlled rumor networks "sibs,"
slang for sibilance" Thus, the .very operators: of deception networks conceived of
their work as the production=of background `.`hissing."
Although the artful' deception operator may blend his disinformation products
with the natural .background; noise, our failure to segregate the spurious signals-
what I call sprignals:. from: the : background noise leads us to underestimate the
importance of these..sP. n"gnals "arid denies us the opportunity to. design counter-
deception systems.
A more helpful model 'of erroneous intelligence would segregate genuine sig-
nals, unintended clues of-futtire events; sprignals, intended clues to nonexistent
capabilities., or intentions; and noise, the background data and random confusion
inherent in the intelligence picture. Whaley's revised Codeword BARBAROSSA
,_comparative -case. studies in Stratagem.
ui .. Jvtany,oi Lne 1151gnts to vronistetter s mas%errui .ream Jiarporszuay are stiff
pertinen~ the problems of compartmentation, of overzealous security, of varying
perspectives, of wishful thinkin9;0 f tangled jurisdictions; of sheer overwork, and of
;strategic deception, but they should not deflect -attention from the searchfor coun-
termeasures to what appears to be the nrost`d"egrading factor of all, stratagem, the'
the' inbient. background noise. These contributions to error becloud therole'of
 Approved For Release 2007/04/17: CIA-RDP83M00914R00230?0090004-8
t -
22 .
subject to deception. Minimizing depehdeice:.uponx timely warning is not enough;
sensible planning requires sensitivity to the-possibilities of adversary deception, and
One may look with skepticism upon a "rational model" of decisionmaking"
while acknowledging that amidst the noise, amidst the bureaucratic politics, amidst
the irrational and the confusing, there lurks the hand of rational deception-admit-
tedly a hostile hand, taking advantage of irrationality, of preconception, of confu-
sion. By misdirecting the apparent thrust of policy, deception can facilitat the
misdirection of resources; by enlarging or prolonging the appearance of-raiiltiple
options, deception can force the rational adversary and probably the,scared, irra-
tional adversary to spread his resources thin.
In many cases there maybe no perfect countermeasure, but surely?we.ought
avoid surprise in more than 4 percent of the major deception cases O It is erely .:._
contended that a modest investment in a counter-deception system would be highly Z
cost-effective should such a system manage to counteract even a minority of decep-
tion operations aimed at decisionmakers through the eyesand ears of U.S. intelli-
gence. Even a low probability timel forewarning :'pnay:iscqurage the prudent;
those who are bent upon warning-free attack may yetber issuadedby the production
of timely strategic warning; and those who face " enhanced::prospects of alerted,
protected retaliatory forces may be further discourager finmhugh-cost procurement
of weapons to attack alert-dependent retaliatory system
Althou h enhancement of warning reliability is a valua~,e.ob}ective, the record.
of past warning failures confirms the wisdom-:of procuring weapons systems that are
not warning-dependent, even at high::cost. But even those systems designed to sur-
vive a surprise attack may depend. upon targeting'iniingence, which may also be
adoption of prudent countermeasures.
before turning to them, we should attach three caveats to this pessimistic prediction
prise.a . realarming. The implications for arms control negotiations are serious, but
Professor W ialey'sJempirical data on deception and surprise in 20th century
warfare lead us to the unpleasant conclusion, qualified by the predictive limits of
Bayes' Theorem," that.itj5.possible to deceive the intelligence services of the major
powers in,most strategic military initiatives; and in the most critical cases, where
one-shot-deception "assets" are exnendead_ the nrnhahilitiac of hivh nt_Pncity cnr+-
SPECULATIONS--.ON THE- VULNERABILITY OF ARMS CONTROL.
AGREEMENTS
PPtofessor Whaley's-. data.s" Third, it is not. clear that the set of 114 cases in the -
;.:constitute such :- ammaltered, condition .and invalidate the Bayesian utilization of.
:evidence from historical.cases may not~validly predictfuture.cases where conditions
inilitai thpabilities:may?suffice for many . arms, control:purposes. Second, empirical
o-go; ti6*e!; place; strength; style) as an intelligence failure, accurate predictions of
--first,';although the Whaley data count any one of five modes-of surprise (go,
?:of intelligence`capabilities.
y of causality are altered; development of effective- counter deception systems would
L,,,,+the same_uaiverse; some of the arms control verification cases may not involve
 Approved For Release 2007/04/17: CIA-RDP83M00914R002300090004-8.
Whaley studied are not always central to problems of planning, negotiating, and
tem, but the generic. problem of 'deception remains: Reassurance costs are high,
opportunities for miscalculation are recurrent. -
How vulnerable are arias control agreements to strategic deception? In candor,
we should admit that we do:not know the answer. Fortunately, the vulnerabilities
assessments of enemy intentions and resource allocations, which have led to such
a poor predictive record in warfare. Thus, deception of verification systems may not
be an insuperable obstacle to arms-control agreements, though it should worry any
thoughtful planner.
After studying the pernicious effects of deception upon predictive per ?orniarce,
one may conclude that the greatest test of intelligence systems is distorted.light, not.
.darkness."3 For the int sligence analyst ignorance isbliss, at least in contrast to the
deceptive reinforcement of preconceptions."' The "shotgun" approach to `iritelli
gence collection will yield a few clues sooner or later, probably sooner, but=if these.;
clues are misdirected, the accumulation of dissonant information may not stimulate:="`
attitude change and production of sufficiently unequivalent;:timely warning."'
Enthusiasts of arms control agreements sometimespose-hy-pothetical verifica-
tion "challenges" of the pure "darkness" variety; naturally, there is sufficiently
penetrating light to conclude that mere security or-pa s5iive camouflage cannot hide
most major weapons developments. Jeremy J. Stop i his?article on ABM detec-
tion, "Can the Communists Deceive Us?" claims thattleycannot He predicts that
the Soviets could not develop and deploy a new generation ABbi system without U.S.
intelligence having timely warning well before the ABM system became operation-
al. Stone cites development of multi-sensor, multi-spectrum reconnaissance systems
as proof that camouflage would notbean'efectve.shield. And clandestine agent.
reports are seen as another mode of advance' s-aEram
But what of an ABM deployment'.thatconsistsofthe selective improvement of
a sophisticated air defense system, such =has the TALINN Line;" ABM deployments
along the Sino-Soviet border with residual capabilities against U.S. strategic mis-
siles; or experimental stations used in space; exploration? Can we be sure that we
would not misrceive the.available technical intelligence data, partly as a result
of clever leaks, indiscretions;: and camouflage? Should we regard the bulk of our
clandestine agent channels as ?alternative modes of warning or as reinforcing chan-
nels of deception?. Peihaps'with'expensive ferret satellite investments we have
reassured ourselves about' ti a present limitations of the TALINN air defense Sys-
living *ith'arms control agreements. For formal or tacit arms limitations, assess-
ment of the present military capabilities and technological resources of the major
existing conditions is the subject of intensive search; deception planners may
powers may 'provide much of the required intelligence. Here, at least, a set of
resent flse-pict bt thdlilii bld
aures,ue uneryng reates maye gimpse from varying
-ln 'contrast, the wartime intelligence~predictions in-Stratagem involved five parame-
ters, ibtr'of which did -not ~exist',at>the'inoment-ofprediction. Whaley's categories
of surprises included the=future'intentions of the adversary, attack or no attack; a
future time of attack; a place of attack, perhaps undetermined and as yet unassault-
ed; and the style of future operations. The strength of forces, Whaley's fifth parame-
or it'may eve ini a subsequent. stage; buttthere,isanobjective reality, and viewed
in`enoiigh lights this reality will probably reveal tliecontradictions within deceits.'s'
 . Approved For Release 2007/04/17.: CIA-RDP83M00914R002300090004-8
r- - -s
244
... , NOTES T4--SECTIQN-:Ili., .... r_.
reality to serve as a yardstick for evaluation, these estimates of the future
as vulnerable to-strategic deception as those in W h a l e y ' s wartime cases. Is substan=;:
tial arms control verification the panacea? Hardly, for verification implies that there
is an objective reality serving as a standard for authentication Verification will help
to confirm the realities of the present, but the main* vulnerabilities of intelligence
services lie in their predictions of the future. For future projections there can be no
verification; inspection, or licit intelligence collection, ;may;:provide clues to the
future, but these clues may be more susceptible to deception than those that con-
tradict evidence of the physical presence. On-site inspection, if properly safeguarded
against deceptive "plants,"_may reduce the susceptibility to a'`.:`-'
Lion, but it cannot
eliminate deceitful projections of the unverifiable.
In conclusion, we do not have an ;precise`~mean of gauging the vulnerability
of strategic arms limitations agreements toy deceptioave reason to hope that
the predictive records of the major intelligence services will be better than their
performance in 20th century warfare. We Ihave"reason to fear that in the most.
important cases, particularly those involvi.ng.iinverifiable future choices, there is a
high probability that strategic deception could yield surprise and advantage. We
have reason to believe that'.various arms limitation programs could be designed so
as to reduce the importance of deceptively induced surprise. And we have reason to
believe that some:strategic:postures (e.g., those involving substaptial advantage to
the initiator of.afirst strike}would: accentuate the importance and likelihood of
deceptively induced surprise.' Rather : than emphasize "verification" of the national
intelligence or international inspection varieties, we should concentrate upon devel-
oping counter-deception methodologies and systems.
Strategic deception;'in*modest doses, has been associated with distressingly
high incidences of surprise` in war. High-intensity deception has been associated
with.: gb.-mtensity surprise and an incidence of surprise of 96-F percent. Given the
.4:lmultiplicity of;challenges in intelligence forecasting, deception is not the main
ter, involved estimates of future choices among alternative resource allocations.
Thus. Whaley's cases of surprise-involve many of the hardest predictive problems.
Fortunately, the intelligence requirements of strategic arms negotiations are
weighted with appraisals of existing realities. These are not impossible but difficult
to falsify; deceptions that succeed for a time may-be discovered in the comparative
evaluations of perceived realities ?O Present military capabilities, technological re-
sources, even political interests and incentives may be gauged. But future: capabili
ties, goals, incentives, and resources must be judged as well, and .here -the.,poor
predictive record of the Stratagem cases should be borne in min -with no:
in .ent to accurate war warning; but methods for the reduction ofvulnerability
pest .'.`track record" of warning reliability should stimulate a quest for better gre-
-"Jb: adversary ,deception may contribute to enhanced warning reliability. And the
::stack:warning methodologies.
1. Bittman's The Deception Gameis the first public exposition ofa covert politi-
cal deception system at work. .
 Approved For Releas.e.2607/04/17: CIA-RDP83M00914R002300090004-8
i 2. Ewen Edward Samuel Montagu, The Man Who Never Was. The Story of.
Operation Mincemeat"(London: Evans, 1953).
3. Chicken feed is false information, or true but relatively harmless informa-
tion.
4. Neither Frank Owen's book, The Eddie Chapman Story, nor Eddie ."Chap-.
man's The Real Eddie Chapman Story(London: Library 33, 1966), more than
hints at the deceptive exploitation involved. Only the preface --i.n?' t#ie.latter
(London: Barker, 1966), esp. ch. 8; "Prosper," pp:-':l 95-214; ch. 9, "Pacts with
the Gestapo," pp. 215-243; ch. 11, "The Balance of (reason ".Pp. 280-306; ch.
14, "Dutch Tragedy," pp. 390.407; ch. 15, "The Broker-Code, " pp. 408.432; ch.
16, "The Great Impersonation," pp. 433458; cli _17 'tlnDefeat Defiance," pp.
459-486; ch. 18, "Dutch Epilogue," pp. 487.502. Iui:R. D Foot,. SOE in France
(London: MM, 1966), esp. pp. 105-110, 307-309, 312=314;:321-322, 326-349,
work is of much assistance.
5. R. V. Jones, "The Theory of Practical Joking Its Relevanceto Physics,"
Bulletin of the Institute of Physics (June -1957), p. 195. Ainrom H. Katz = ' v
brought this. article to my attention.
6. Operation NORTH POLE: See, among'others, E , . Cookridge,. Inside SOE
350-351, 368, 375-376, 380,'386-388. Philippe Ganier-Raymond, The Tangled
7. 'Jacques Mordal, "La tenebreuse a$'aire,de'Gleiwitz," Miroir de l'histoire
.(Munchen: Stutz, 1950).
Pole (London: Kimber, 1953). Joseph:Bchrieider, Des:.War dos Englandspiel
Web (New York: Pantheon, 19683 Hei .iian: J.Giskes, London Calling North
(October 1962), pp. 469-478, and La gnerze a commence en Pologne (Paris:
faked incident -at the U.S. naval base at Guantanamo Bay as an excuse for
a U.S.-supported invasion of Cuba. Transcript of Interview, George A. Smath-
guerre (Paris: Arthaiid,:1961); E. Aleksander, Tv Radio Gliwice (Warszawa:
Wyclawan Min...Oborony, 1967): Senator George A. Smathers suggested to
President-elect. Kennedyy * similar deception operation: the mounting of a
Presses de la Cite; 1968). Gunter PeisNaujocks, 11omme qui declencha la
ers, Kennedy. Oril.#istory Project (March 31,1964), cited in Henry Raymont,
Denne ie "Le sosie de 'Monty,"' Historia (February 1960),. pp. 201-204; Ste-
8. A Cecil Hampshire,-The Phantom Meet (London- Kimber, 1960), P. 208.
-9:. X. -E. Clifton James, I Was Monty's Double (London: Rider 1954)? Bernard
"Kennedy Arcliives.IIluminate Cuba Policy," The New York 7ymes (Aug. 17,
1970), pp. 1, 16.
1961),,*--! %Was Monty's Double Once Removed," pp. 158-173. Various associat-
'plienVatts, Moonlight on a Lake in Bond Street (London: The Bodley Head,
ed deception operations, real and. imagined
are described in the unreliable -
,
;" work by'Gilles Perrault, The Secrets ofD-Day(Boston: Little Brown, 1965).
icier -'Sevin, "Operation Scherborn," MilitaryReview, 46 (March 1966), pp.
43e-
11. David E. Walker, Lunch with a Stranger(London: Wingate, 1957, New York: . .
Norton, 1957), esp. pp. 152-155,160-223. See also the "Morale Operations" of
OSS operative Elizabeth P. MacDonald, UndercverGirl(Nev-York Macmil-
lan,1947), esp. pp. 2, 7; 8,'and 20. Some peacetime examples, of disinformation
are found in: U
S
Senate
Committ
on th
J
di
iar
In
r
.
.
,
ee
u
e
c
y,
te
nal Security
Subcommittee, Communist Forgeries, Hearings, Testimony of Richard
 Approved For Release. 2007/04/17 CIA-RDP83M00914R002300090004-8
Helms, A.ssistant Director, Central Intelligence Agency (Washington, D.C.:
1961). p. 121; Thomas Whiteside, An Agent in Place--The Wennerstrom Af=
fair(New York: Viking Press, 1966), pp. 144.147, as "Annals of Espionage, An
Agent in. Place-III," The New Yorker (April 9, 1966), pp. 149-150;: Peter
Sager, Moskaus Hand in Indien (Bern: Schwe-izerisches Ost-Institut,1966) as
Moscow's Hand is India (Bern: Swiss Eastern Institute, 1966; Bombay: Lal-
vani Pub. Co., 1967); and Ladislav Bittman, The Deception Game, (Syracuse,
New York: Syracuse University Research Corporation 1972):::::::::::_ 12. Peter Wykeham, Rghter Command: A Study ofAirDefense, .1914-1%O"n-.
don: Putnam, 1M), esp. pp.164-165,167,168,171,175,177,227-228, 239-240,.'.'.-
# IA API .
13,. N. Gordeyev, "Operational Camouflage in Naval Landing Operations," [in
Russian] Voyenno-Istoricheskiy zhuraal Moscow. IApril 1969), pp. 41-51;
trans. JPRS No. 48,346 (July 3, 1969), p. 32.,-
14. The best single study of electronic deception; in World,War U focuses upon
British accomplishments: Alfred Price, Iasblumeats of'Darkness (London:
"
Kunoer, 1967), P- 254, esp. ch. 1,
Doctoring'-the-Beam. ," pp. 19-51; ch. 5,
"Doubts, and Decisions," pp. 111-130; ch. 6, "The Rude--Awakening," pp. 131-
149; ch. 7, "Gomonah and After,'.': pp. 151-177; ch. 8, "The Climax," pp. 179-
"
198; ch. 9,
In Support of Invasion'. .pp.199 211; ch. 10; ?"The. Last Round,"
pp. 213-247, including reference to the foUowing code names: ASPIRIN, BEN-
JAMIN, BROMIDE, CIGAR,-DOMINO DLTPPE;FILBERT, HEADACHE,
"Prague Invasion:Vanguard Used Aeroflot; Russian Civil Carrier Flew Two
Secret Police Teams into Ruzyne Airport to Spearhead Air Assault, Seize
.:..Czech Party Headquarters," Aviation Week & Space Technology (Sept. 30,
JOSTLE IV, MADREL, MOONSHINE,.TINSE~:TITANIC, and WINDOW;
also STARFISH and "Col. Turnei's:I)epartrnent " See also: Paul Brickhill,
The Dam Busters (London: Evans; .1951), pp. 190-196; David Irving, The De-
struction of Dresden: (London: Kimber,' 1963; R. V. Jones, "Lees Knowles
Lecture Ill Comzrandkand IntelligenceCambridge University, 1968, pp. 23,
24-26,29; James DeRoeher and William Huse, "EW Support.to Tactical Cover
and Deception;" Electronic Warfare 3 (Convention Issue,1971), pp. 16-17+.
15. Sedm Prarskych Dnu 21.27 srpen.1968 The Seven Days in Prague (Praha:
Inst. of History, Czech Academy of Science, 1968), abridged and ed. Robert
Littell, TWO-'Czech.Black Book (New York: Praeger, 1969). Donald E. Fink,
:1968);-pp.16-18; Rittman, The Deception Game, pp. 167-215. The best public
kehaveyatamt[Things''as They Are~};(Tel-Aviv: Am Has"eoher,1965). E =Hasin
;16 . Yehoshna Arrel, Ha kenanyah (reI .Aviv:1965). David-Ben-Gurion, Devariin
y and political Change, (New York, Praeger, 1973).
Surprise," pp. 31-49, revised edition in Gregory Henderson (ed.), Public Di-
survey:. of this case is Barton S. Whaley, Public Diplomacy Aspects of the
Soviet Invasion of Czechoslova.kla in 1968(Medford, Mass.: Fletcher School of
Law and Diplomacy, Dec. 1, 1969), ch. 5, "Invasion! Deception as the Key to
and D. Horowitz, Ha Parashah,'[The Affair] (Tel-Aviv: Am-Hasepher, 1961).
Hagi Eshed, -"Ha-Parashah" J-[-The Affair] Ha-Aretz`(February -19, 1965);
Kenneth Love, Suez The T vice-Foctght. War(New York: McGraw-Hil1,1969)
pp-48, 52, 73-80; Amos Perlmutter, Military and Politics in Israel (London:
_,Frank Cass, 1969), "The Era of Consolidation: The Lavon Affair, pp. 80.91.
 Approved For Release 2007/04/17: CIA-R DP83M00914R002300090004-8
.102-104; Howard M. Sachar, From the EndsoftbeEarth: The Peoples of Israel
(New York: World, 1964); J. L. Talmon and Ze'er Katz, "The Lavon Affair-
Israel Democracy at the Crossroads," New Outlook-, 4 (March-April 1961), pp.
23-32.
17. In re Burma, see Stanley Karnow, "Record of CIA in Southeast Asia -Places
U.S. Name in Disrepute," Providence Journal (Sept. 7, 1965); Congressional
Record 111. pt, 17 (Sept. 13, 1967), 23486-23487; as "Espionage; Attempts in
the 1950's Recalled; U.S. Image in Southeast Asia Suffers From -Cluiasy.Ir -.
,
Record, 111, pt. 18 (Sept. 20,1965),24421-24422. In re Indochina, see Bernard
trigues of Agents," The Washineon Post, (Sept. 7
1965), iii''Congressional
B. Fall, Hell ra a Very Small Place. TheSeige ofDienBien Phu (Philadelphia"',
& New York: Lippincott, 1966), pp. 41-43, .49, 76,:228; '317-318, 342. David
Walker, Lunch With a Stranger, mentions the problem of deceptive feedback
interfering with one's own intelligence estimates; in SOE :vernacular the
problem of "sib" comebacks.
19. William Eliscu, Count Five and Die: How AlliedResistarice. Leaders Were
Deliberately Sactzf ced to Gestapo Torture--The True Story of Our D-Day
Deception Plan, "Operation STAMPEDE"(New York: Ballantine,1958). The
official SOE historian, M.RD. Footas less than candid when he observes, in
SOEin France (London: HMSO. 1966j'p 464:'" ere is nothing in the files
to show that this is anything but fiction " Professor-Foot tells elsewhere (1966
ed., p. 450) of the mysterious loss :of:SOE deception files, and explains their
...unavailability (1966 ed., p. 451) for his :history. Thus, his evaluation is built
on a shaky, undisclosed foundation.
20. Gilles Perrault, TeRed.Orchestra (NewYork: Simon & Schuster, 1969)) pp.
131-151,156,187-188,.198, 205, 210. This unreliable account is the subject of
a critical review by Christopher Felix, The New York Times Magazine (Au-
gust 3, 1969) :pp. 3, 18.::'-:
21. Antiquity:": S_unn-'I`zu, Sun Tzir',Prng Fa [The Art of War,] ca. 510 B.C.; Sextus
Julius Froritinus, Stratagerala, ca. 90 Al); .Carlet de Ia Rosiere, Stratagernes
de Guerre, 1756-On-Napoleonic deception: David G. Chandler, The Cam-
paigns afNapoleon (New:York, Toronto & London: Macmillan, 1966), pp. 74,
78-84,146-147,151-155,:180-181,186-189,196-199, 214-216, 276, 278; 280-281,
'.289-291, 384,386.387,389-390, 394, 396397,405406, 409-433, 436n, 528.530,
:604605; 711-717.
22. See tbi: iiithoritative study of strategic deception and surprise, Whaley,
Stra*t ig . Whaley's study is, in my opinion, the most significant public work
on the theoretical aspects of intelligence activities published in the 1950s or
--196L)s - Whaley weighs the relative importance of security, camouflage, cover,
(arid deception in enlarging the probability and usefulness of surprise in "The
iincidence of Surprise with 'Deception," pp. 163-169, "`The Causes of Sur-
prise," pp.` 170188; "Tt a Varieties and Intensities of Surprise and Decep-
tion,' pp. 210-223; "The Economics of Stratagem," pp. 232-244. -
The CIA's involvement in the Guatemalan coup d'etat of 1954 has been
generally acknowledged, but the operation's' adoption of a, psychological
deception/communications strategy has been only rarely perceived. See
_~__ 'arnotsu Shibutani, Improvised News? A Sociological Study of Rumor (In
 Approved For Release 2007/04/17: CIA-R DP83.M00914R002300090004-8
(1968), pp. 472473.
33. Ihid;- pp 212.213. Whaley discusses only a few cases of technological surprise.
of deceits. I am indebted for this suggestion to a World War U stratagematist,
Maj..Gen. William A. Harris, USA Met).
Even.though we may assume that the incidence of deception in technological
surzse is well less than tsx percent, we would be foolish to dismiss the role
''cited in Note la above and,R. V,,,Jones, "Irony as a Phenomenon in Natural
Science and Human AS'airs," Chemistryand Industry(1968),,pp. 472-473; R.
V Jones, ".The.Theory of Practical Joking Its Relevance to Physics," cited
,In, Note 5,,.pp. 196-199. Even the theoretical vulnerability to technological.
deception can have serious ramifications. See Albert L. Latter et aL. A Method
dianapolis & New York Bobbs-Merrill, 1966), p. 194. More recently, Ernst
Halperin has outlined the psychological deception program that stimulated
the coup in Guatemala and the similar program, intended to precipitate a
coup d'etat at the time of the Bay of Pigs invasion of Cuba, April 1961.
Halperin's study is the only one that places the paramilitary landings at
24. Whaley, Stratagem, p. 244.
25. Ibid., p. 146.
tal Hoaxing," pp. 178-193.
Playa Giron in perspective, as a diversionary and psychological instrument.
The same study analyzes the psychological deception program of Castro in
195&-1958. See Ernst Halperin, The National Liberation Movemextsan.-Latin
America! (Cambridge, Mass.: Center for International Studies 'Mi:1 June
1969), pp. 13-14, 18-45, 53-67. For a propagandistic rendition of theHa-vana:.,-
coup -that-never-was, see Gunter Schumaker, Operation Pluto. die - r
chichte einerinvasion (East Berlin: Deutscher Militar verlag,1966). Cases of political deception are detailed in Bittman, The:Lleception Game; and in
Curtis D. MacDougall, Hoaxxes(New York: Dover,'1958), ch.13, "Governmen-
26. For other examples of this phenomenon, see R. "V ;1 nil ;."Irony as a Phe-
nomenon in Natural Science and Human Affairs, Chetstry and Industry
29. lbid., pp. 210-212.
28. Ibid., esp. pp. 189-196, 232-24.4:':
z1. whaley, Stratagem, p. 163.
30. ibid., p. 'Y62. Of 33 cases, 28 involved surprise and deception. But only 25 of
the surprises have been associated w thh deception operations, and three
deception operations failed to yield: surprise. Considering that suitable am-
phibious landing areas are generally loth-obvious and few, the achievement
of surprise in 85 percent of landings and in 89 percent of those associated with
deception deserves explanation. A landing force isolated on naval vessels may
be briefed: with- reduced risks: of insecurity and may proc?d by sea with less
risk of detection .than in overland transit But equally insportant is the factor
of planaing;ti e. Amphibious. landings require exact coordination of more
than one military arm. Thus amphibious operations require longer lead times
and provide deception. planners sufficient notice to launch their full arsenal
:32 ' Ibid~'p:*'204.
 Approved For Release 2007/04/17: CIA-R DP83M00914R002300090004-8
37. Barton Whaley, The Bodygua,of`I`ies;: A;delphi Paper (London: Institute of
Strategic Studies, forthcoming, 1973)'esp pt ;:II "'I'be Values of Surprise"
38. Ibid.
sonahty, Ph. D. dissertation, University of Miami, 1971:
35. Whaley, Stratagem,. (1969 ed.), pp.. B9, B15, B23.
36. Ibid., p. 6.
of the technological collection systems, and by the credence put on llie; results
of collecting with these systems.... The more successful these ; systetas, the
more it would profit the opposition to instigate large scale deception, botli`as
a defensive measure and as a possibility for misleading and.planting misinfor- -
mation." F. L. Schaf, Jr., "The Evolution of Modern Strategic Intelligence,'.: ,
Thesis, U.S. Army War College, (Carlisle Barracks, Pa..,May 1965), MS, p. 679. '.*;f'
34. See Arnold M. Ludwig, The Importance ofLying.(Springfield, Ill : Thomas,
1965); A. George Gitter, "Hypocrisy as a Way of Life," Ph.D. dissertation,
American University, 1963; Hugh Hartshorne. and Mark A. May, Studies in
Deceit (New York: MacMillan, 1930). Low. self`-esteens:is :correlated with a
self-deceptive tendency in a study of cheating~lin ~nr tic ~rMrc~ =;+I Mir f. r?Pf_ flip- PnP_mV &m;:^"
TESTING FOR''SIGNAIS:AND.SPRIGNAIS
The foregoing suggests that it may be considerably easier to collect.sprignals
Actually a lot of fairly obvious things_jin a deception prods o] were really
a cover for one's own "unwitting'. :people :who were not to know that other -
more complex and more "classified" operations were going on to really get
to the enemy command decisions.--. his is not:to.gamsay that routine cover
and deception operations of this `sort were not :done to connect with all enemy intelligence systems."
Some sprignals'may beself-identifying; if so, we may search for confirming data
or channels that may yield already anticipated sprignals at some later date. A group
that is sufficiently lucky or. successful in penetrating foreign deception systems--
will 1e' arn'of deception efforts from "agents in place," or from modes of collection "by
technical means." A somewhat less well-established group will work out their in-
complete jigsaw puzzles until some of the pieces fall into place inferentially. They
>: will debrief defecting stratagematists with unusual care, reconstruct the sources of
sources carefully.. ._, ~. _...
insecui ty that may have been turned against them And they'willl watch those
rod security experts for more`pessiiaiistiC~eV luatifrn""s thatidentif `those sourceso
sprignal'identification missions of the counterintelligence community. They will
past patterns of deception. A politician who deceives his compatriots (even in pursuit
The incidence of signals and sprignals_'rnaybe gauged; iripart; from analysis of
of "national" objectives abroad) may not enhance his estieen: 'It should not be.
than to separate them frozassociated signals and noise.
 Approved For Release 2007/04/17: CIA-RDP83M00914R002300090004-8
. 37
surprising that most of the literal statements of leading statesmen are at least
technically truthful, -or at least so intended. The implications of tendentious -re-
marks may be far less reliable--as with the implications of lshrushchev's boasts of
growing Soviet missile production in 1957-1959.12 And the literal or implied mean-
ing of "background" statements not linked to a source whose credibility. is highly
valued may be of lessened reliability. So too with official press statements, -.a? with
the indirect TASSdemal of September 11, 1962 that nuclear missiles wouldbe-sited
in Cuba.13 ';....
The careful analyst may wish to distinguish ' - tat
the past reliabilityoi'literal'stat~:
ments, of inferences from tendentious remarks ripe for the plucking, of `back-::
ground" interpretations, and official but anonymous press releases. One analysisaf
perceived mendacity in and around Washington, D.C. is consistent with observations
of foreign statesmen. Although infrequently finding "the flat lies. .and petty mis.
truths," Anthony hake reports: ;' -
There have been still more misleading statements-not'quite lies, but
partial revelations of the truth deliberately designed tolbol the public into
.--___but not all of those available, and having recognized that this collection is cluttered
If the processes of stratagemic analysis are coordinated with the aid of comput-
ers, it may be possible-to-geaerate'data batiks ad ` analyiie specialists familiar with
the styles of unseen adversaries. Having collected and-identified a mass of sprignals
in government statements, and when. it is fooled the result=isthe same loss
believing what the government wants it to believe. Theseni st.be considered
the functional equivalent of lies.... The public does not read the fine print
as-an autotnatic:standard against which to test new sources or new data.
ulleraliilities;Thus, traditionally reliable sources and data cannot be segregated
The counter-deception analyst eara'fo'.'d 'ereitiate the literal from the
implicit, the flat from the ambivalent 'lie;:.the deceived from the deceiving spokes-
man." Further, by cross-checking deep penetrating technological sensors (by the
"inconogru.ity.testing" method to be discussed shortly) it may be possible to establish.
with high confidence that a source or channel of data is indeed a stratagemic
producer and not a sourcie4signals or noise: if. the exotic sources of the past have
been the favorites of deception. planners precisely because they were the ones that
received top-level attention,: these "special" sources will not be adopted as arbitrary
standards against .which potential..-sprignals will be tested.
It is. equally;impoi It to guard against the temptation of resorting to a test
standard derived-from:."traditionally reliable sources." Over extended contests be-
tween intelligence and'counter-intelligence adversaries, both sides tend to develop
an understanding of what an: adversary views as reliable, and deception planners
target those channels. Data that appear "reliable" are likely to conform to precon-
ceptions.:pP::shared expectations, and deception planners like to reinforce these
parating_..truth from deception and noise involves more than a comparison of
-assume:tha# ail channels of information are unreliable in varying degrees and to
coed i de -th hole of chance in strategic forecasting. By conceding the hazards of the
ritelligence business, one -is more likely to embark upon `an=~bngoiii
siftin
nd
g
g a
resiiatin effort: identi -
g fying;tentative-subsets of signals ar