CIA MEMORANDUM ON THE TALLINN SYSTEM

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP85G00105R000100100003-0
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 4
• Document Creation Date: November 11, 2016
• Document Release Date: July 17, 1998
• Sequence Number: 3
• Publication Date: October 10, 1967
• Content Type: MF

Approved For Release 1999/09/25: CIA-RDP85GO0105R000100100003-0
10 October 1967
MEMORANDUM FOR: Director, Strategic Research
SUBJECT CIA Memorandum on the Tallinn System
REFERENCE Director, Strategic Research memorandum,
same subject, dated 25 September 1967
1. The points made in paragraphs 3-5 in your memorandum
to me are well taken. I propose that we do instruct our staffs
to use their judgment about mentioning ONE's participation in
CIA. memos which deal with the subject matter of important NIEs.
I think we are in complete agreement that measures should be
taken to make it impossible for readers of such documents to
confuse them with a USIB Memorandum to Holders or, for that
matter, with an ONE memo which might seem unilaterally to
amend a community judgment as expressed in an NIE.
2. We do, of course, wish to continue the present system
of coordinating memoranda. Let's not however follow the
procedure used with respect to the memorandum on the Tallinn
System. Our staff saw only the summary. We think it important
we be given the opportunity to check all parts of all draft
memoranda.
Approved For Release 1999/09/25: CIA-RDP85GO0105R000100100003-0
 Approved For Release 1999/09/25: CIA-RDP85G00105R000100100003-0
3. On the point in paragraph 3 of your memo -- namely
the Possibility of Conflict Between Pronouncements Made in
Papers Issued by DD/I Components (and those of DD/S&T and even
DD/P for that Matter) and the National Intelligence Estimates,
herewith are some thoughts. You should regard them as something
I have felt deeply about for a long time (and voiced orally on
a number of occasions) rather than something which the particular
memo in question in itself has stimulated me articulate.
a. The NIE is a community enterprise, and an
important one. All analysts in the community ought
to know what NIEs in their area of substantive concern
say. This is not asking very much because a lot of these
analysts are themselves contributors to NIEs, if not
active participants in their production. Since these
documents are submitted by the DCI (the pronouncements
and judgments which they contain are his) analysts in
the Agency should, a fortiori, know their content.
b. In God's-eye NIEs are probably of different
degrees of importance. Anyone, by the crudest rules
of thumb, ought to be able to identify the most
Approved For Release 1999/09/25: CIA-RDP85G00105R000100100003-0
 Approved For Release 1999/09/25: CIA-RDP85G00105R000100100003-0
important. These, obviously, will be the ones which
are most likely to affect decisions regarding the
US defense budget (consider the No. 11, 12, 13 and 14
series) and other aspects of US national security policy
(consider the Vietnam and Arab-Israeli series).
c. No analyst should modify judgments in this
range of NIEs in a document destined to circulate
outside the Agency without knowing exactly what he is
doing.
d. When an analyst perceives good reason to challenge
or alter one of the important judgments of one of these
critical estimates, I would like to think that he would:
(1) invariably call the matter to ONE's attention;
(2) confer earnestly with appropriate members of the
ONE Board and Staff; (3) and make sure, before going
into print, that the document he proposes to circulate
will signal in some way or other the fact that its
authors know what the relevant NIE says and how they
feel the judgments of that NIE should be altered.
Approved For Release 1999/09/25: CIA-RDP85G00105R000100100003-0
 Approved For Release 1999/09/25: CIA-RDP85GO0105R000100100003-0
e. For example, I think analysts should take
special care when producing anything on a subject as
important (and as much a matter of community concern)
as the Tallinn System. In my view, a memorandum like
the one under discussion should make mention, in the
text or in a footnote, of the fact that the matter is
treated in existing estimates, that it will be treated
again (when such treatment is scheduled, as it was in
this case), and indication given to the reader of the
nature and extent of differences, if any, between the
judgments in the memorandum and the most recent
estimate.
25X1A9a
SHERMAN KENT
Chairman
Board of National Estimates
cc: Director, Scientific Intelligence
D/DCI/NIPE
Approved For Release 1999/09/25: CIA-RDP85GO0105R000100100003-0