SEP 15 1977
Dr. Joshua Lederberg

. 1431 PITMAN AVE.
PRECEPTION AND CONCEALMENT PALO ALTO, CALIF. 9430
(This essay introduces "preception’ as a term intended to be | C cog i

broader than 'deception', embracing the whole span of measures
intended to influence the perceptions of the viewer -—- by whatever
means, and whether true or false. In common usage, a precept is a
rule of behavior that a preceptor wishes.to induce in his audience.)

Intelligence analysis resembles scientific research in many ways
— fragmentary data often evoke a range of hypotheses, and in turn
stimulate the collection of more information designed to corroborate
or falsify them. Intelligence entails one unmistakeable difference:
the adversary is witting \*/ , uncooperative and self-consciously

 

\*/ In fact this term is almost unique to intelligence jargon

 

interested in the outcome of your analysis. (The invocation of
‘Murphy's Law’, that nature exercises a malevolent will, is in fact
put down as either a superstition or joke —- so let us borrow the
name Murphy for our adversary.)

Knowing that Murphy lives means that we are constantly on guard,
in the interpretation of intelligence data, that we might be being
deceived. However, it is often difficult to prove that Murphy is
actively fabricating data for our benefit; further it is often argued
that deception cannot be sustained for long periods in peacetime. The
net result, all too often, is that the possibility of deception is
always acknowledged, but perhaps not often enough actively pursued.
Even more important, the moralistic dichotomy -- either Murphy is
lying {the villain!}, or he's not -—- blinds us to a much wider range
of preceptions that Murphy may be exercising, and which then
distinguish the data of intelligence from those of natural science:
namely, that a witting Murphy will be confounding us in relation to
{his perception of} the national security values we attach to our conclusions.

Far from being the unique province of a villainous bureau of
‘disinformation, preception is warp and woof part of all human
“discourse, and above all of interstate relations where mutual
peréeptions are so vital to security and survival.

This nonrandom association of induced noise with values at stake
is the main principle by which preception can be identified, short of
glimpsing into Murphy's soul and intentions. In fact it can be very
‘nearly unconscious on the part of the preceptor —- which of us
does not put his best foot forward? There is then no algorithm by
which preception can be diagnosed, no more than there is a general
system of hypothesis and analysis by which scientific truth can be
mechanically generated from research data. The possible role of
preception is a hypothesis that needs to be considered, presumably
always (in the light of human nature and interest, and tested ad hoc -
along with the rest of the analysis of a given set of observations.

Our basic model is that of a _signal-emitter and an observer.
Everyday experience in-human relations probably has taught us most of

|
the basic lessons of preception; the intellectual block is in applying
the principles of assessing interpersonal discourse to the realm of
technical collection systems. Here we do not think so readily of the
signals as communications; and we are liable to think of technical
intelligence as if it were the observation of Nature. If, like Nature,
Murphy is truly oblivious that anyone is watching, there can be no
preception: the signals emitted will be simple side-effects of his
other activities. However, few if any state: activities can be relied
upon to follow this model.

Information-security is also a form of inverted
preception: signals are selectively guarded, and others not, in
accordance to values of the observed/observer. This shielding then
falls on an axis: 7

Total Selective _ ' Total
Candor Persuasion Revelation Secrecy

 

the poles of which are never seen in the real world.

A separate axis has to do with fabrication — be it of
communications signals, objects, or events [e.g. ‘theatre']. The
dividing line that distinguishes lying from selective revelation is
hard to define — e.g., when the contextual details of: events are under discretionary
control. It is a truism that embarrassing news :‘ories are leaked on
weekends: is that a fabrication?

In analysis, the focus should not be on deception (lying), but
on preception —- i.e., all the measures that Murphy might be using in
his self-consciousness that we are watching. Perhaps the commonest
strategem is for a human agent to tell the truth
with the design of enhancing his credibility or his naivete, “ ve
which is preceptive in so far as it is not «  pappaver
his standard behavior whilst unobserved. . Aue ee

Other axes of the means of preception concern the use of
overt vs. covert channels. The latter implies interception suspected
by Murphy: this is then turned into a means of controlled communication
to the observer. In this context, it may even be more credible; there
are historic examples where TRUTHS have been intentionally Conveyed
by covert channels, open displays having been skeptically dismissed.
Where large investments have been made to penetrate such channels,
there is a dangerous bureaucratic incentive to minimize Murphy's bo oh,
chances of turning them for the whole range of preceptive activities. © © ew east
It is just here that the truth/deception dichotomy may be the most 37). > We we
costly. The chief problem that Murphy faces is in getting feedback —
about his efforts: but we can hardly afford to be too complacent about ~
his ability to penetrate our own channels both in the . ce
execution of preceptions and in providing informational support for
his larger frame. Damage assessments of discovered breaches should
include this agenda.

 

TC

Historically, human agents have been notorious vehicles of
<preception; and it is doubtless true that these games are much easier

—)-
to play within the framework of the intelligence/ counter~intelligence
bureaucracies, than when they involve the recruitment of other economic
or even military actors. This should alert us to be most suspicious

of preceptive games when intelligence, cover or concealment.is already
evident, albeit for. other manifest ends.

 

The discussion so far has focussed.on MEANS of preception.
Can we catalog the GOALS? Probably not exhaustively, since this would
embrace the whole framework of the belief-structures of state subsystems.
We can factor out psywar-operations, insofar as these are directed to —
public morale rather than decision-makers: privy to special sources of
information; but the basic principles, are not.greatly different. But
it may be easier to fool some of the people more of the time; preceptive
communications directed to military and political leaders will be more
complex and sophisticated.
A rough taxonomy of goals would possibly include:
surprise
conditioning
sensitization .
diversion, distraction and saturation
probes (evidentiary tracers) |
(a collection that does show some overlap).

Strategic military surprise must command the highest priority
of our concerns.
The works by Barton Whaley and Anthony Cave Brown give abundant examples.
Hardly any of Hitler's successes were not facilitated by masterful
manipulations of the assessments of his rivals; these are not well captured
by the term 'deception'. More generally, military surprise has always —
had to rely on tight security and diversionary tactics as well as overtly
fabricated deceptions. That it is far easier to run such a deception in
-wartime, and on a short time frame, is readily stipulated. None of the |
other speculations about uses of preception are intended to detract from
the superordinate importance of this one.

For long-range applications, however, the CONDITIONING of our
belief systems in various ways deserves priority of attention. In fact,
beliefs that we cannot be surprised [e.g. on the NATO front]
must be the most suspect — and
deserves analysis in terms of the possible role of preceptive games in
having engendered that view.

In general, conditioning is a manipulation of our perceptual
system, of our way of interpreting data, rather than the implantation
of a concrete belief. It is then an investment for future exploitation.

The repeated association of a given signature with a target object
is, for example, likely to influence how we count such objects, or how
we interpret such signatures in future. The most wearing games would
involve ringing the alarms repeatedly —- as the Arabs did prior to the
1973 war — to the point where it is futile to respond to each alarm
(especially if ringing the bell is cheaper than sending out the fire-
trucks, and the alarmer can light the match at his discretion).

This form of desensitization is a saturation of material resources
available to respond to alarms; others are a saturation of attention.

-3-
We can be desensitized in other Oe less specific
wayS, @.g., to sheer novelty. An enigma seen many times ends up in a
mental and bureaucratic pigeonhole. They might be multiplied for no other
reason than a general saturation of collection/analysis; and of course
they can be emplaced at places and times to create specific diversions
from more significant targets. These games can also be used to
establish legal precedents to legitimize future harassments of intelligence
activity.
It is characteristic of active conditioning that the signals are

INTENDED to be seen. At the first order, they may seem almost too conspicuous.

Non-display is also a conditioning device, to induce complacency:
‘but we call that security (Murphy will wait for a surprise attack to deploy a

previously concealed weapon). And the same principle of waiting for high

stakes will surely be thought of for the preceptive games themselves.

Another branch of conditioning is senzitization. Exaggerated
self-skepticism about the reliability of one's intelligence could multiply
the costs and deflate the value of the product. It then pays to inject
manifestly spurious information into covert channels, not to deceive, but
to sensitize the observer into mistrusting his interceptions. {The main
motive not to do this is of course to sustain the image of naivete for a
future deception.} This gambit will be most readily thought of as a
recovery operation after a security breach; but it is akin to keeping
the alarms perpetually on red, or to jamming as an alternative to encryption
for data security.

In peacetime, preceptions are thus oriented to counter-intelligence,
but can also be used to divert our military, political and technical
resources to various wild goose chases. {We would then
be well-advised to reflect on the history of our adventures, to discern
which of them may have been seeded by the other side. A possibly trivial
example is military interest in parapsychology, though it is difficult to
-be sure who is taking the initiative in focussing attention on this field.
Others may be more sinister in terms of the resources that are drained. I
would also admit that inadvertent (?) mispreceptions, like the supposed
Nazi investments in biological and in nuclear weapons during WWII, may have
huge consequences.} The hazard of playing this game is also well
illustrated by Khrushchev's missile-gap preception of the 1950's —
that we may so vehemently overreact as to generate new threats.

Finally, preceptions thight be planted as probes of our policy and

intelligence processes. Murphy can better trace them if he has made

controlled inputs that could only have come from his sources. There are
-well known historic examples of the use of this gambit to trap suspected
espionage agents, but far more subtle applications are also possible.
Questions like: with whom do we share our intelligence information? What
is our response time to alarms of various kinds? What channels are invoked?
Can cleartext be planted to probe our cryptions? —

What is the discriminating power of our collection systems? What fraction
of sensitive sites are targetted —- this is analogous to counting wildlife
by tagging and releasing samples? All these could be probed by. planted
preceptions.

—U-
Sources of vulnerability.
Human psychology.
Bureaucratic problems.

Asymmetries.

With rare exceptions, effective preception involves a collaboration
of the belief-system of the viewer, viz. some degree of self-delusion. The
roots of rigid thinking in human psychology have been reviewed in two recent
remarkable books (Jervis, Janis & Mann) which should be on every policy-maker's
and analyst's desk. They go far beyond the principles of wishful thinking and
personal self-interest; but I leave further elaboration to those monographs.

Other speakers at the conference have mentioned some of the bureaucratic
problems of attempting to manage the risks of preception. The difficulties of
communicating sporadic suspicions, and the personal hazards of being labelled
Cassandra deserved special attention. They. are connected with the larger problem
of how we deal with contingencies at the wings of the probability distribution,
unlikely (does that mean p < 0.10 ?) ones that may have very high stakes. The
Cassandra by definition will be wrong most of the time; how sustain his morale
and our own balance about false alarms nevertheless? To give another illustration,
does ACCIDENTAL WAR receive the analytical attention and the procedural safeguards
that it deserves in proportion to the relative odds that it may be the trigger?
What is called for is a systematic search for the ways in which our
bureaucratic organization may impede clearsighted responses to the whole spectrum
of threats. One point that I would emphasize is that it must be very difficult
for an organization that has had to struggle for and defend very large investments
in technical collection systems to be willing to entertain the hypothesis that
these can be turned into channels of disadvantageous preception. If I were the
President, I would want to be sure that such an organization was leaning over
backwards to forfend its potential bias of commitment, or that another one
could study this problem from a detached standpoint.

Finally, we all recognize that Murphy may have many advantages in -
the peacetime conduct of preception, particularly in his capacities for
security and for the recruitment of a wide range of witting or unwitting
coparticipants. For a realistic assessment of this asymmetry, it would be
important to collect the instances of evident involvement of military and
civilian cooperators outside the central organization. Bittman's book implied
very heavy involvement of all scientific activities, and a realistic understanding
of measures that go beyond overt security is important not only for intelligence
but for the justifications of various media of scientific and technological exchange.

nam nee re an oe ae

A note about the future role of preception.

Our particular Murphy was badly bruised by the Cuban Missile Crisis,
and I would be surprised if that experience does not dominate his thinking about
the use of these games. When I wear a red hat, I find myself thinking that I am
going to invest a lot of effort in testing my preceptive devices before I rely
upon them again in a high-stakes adventure.

ee ney naan ene eee
References:

Jervis,R. Perception and Misperception in International Politics. 1976
Princeton Univ. Press.

Janis,I.L. & Mann,I.L. Decision Making: a psychological analysis of
conflict, choice and commitment. 1977. Free Press.

Whaley,Barton Strategem: Deception and Surprise in War.
1969. Center for International Studies, M.1I.T.

Brown, Anthony Cave Bodyguard of Lies. 1975. Harper & Row. —