Office of the Director of National Intelligence Washington, DC 20511 Mr. Steven Aftergood SEP f 9 2014 Federation of American Scientists 1725 DeSales Street, NW, Suite 600 Washington, DC 20036 Reference: FOIA Request DF-2010-00031 Dear Mr. Aftergood: This is in response to your 23 December 2009 email to the Office of the Director of National Intelligence (ODNI) (Enclosure 1), in which you requested, under the Freedom of Information Act (FOIA), copies of two ODNI SHARP (Summer Hard Problem Program) reports from July 2009.
Your request was processed in accordance with the FOIA, 5 U.S.C. 552, as amended. A thorough search of our records and databases located documents responsive to your request.
ODNI reviewed one of the documents (Enclosure 2) and determined that material must be withheld pursuant to the following FOIA exemptions:
- (b)(l), which protects properly classified information under Executive Order 13526, Section 1.4(c);
- (b )(3), which applies to information specifically exempt by statutes, specifically 50 U.S.C. 3024(i), which protects intelligence sources and methods from unauthorized disclosure; and
- (b)(5), which protects privileged interagency or Intra-Agency information. If you wish to appeal ODNI' s determination on this request, please explain the basis of your appeal and forward to the address below within 45 days of the date of this letter.
Office of the Director of National Intelligence Information Management Office Washington D.C. 20511 The Department of Energy (DoE) also conducted a review on the material and determined that the document contained Restricted Data (RD), which has been redacted under Title 5, U.S.C. 552(b)(3) as containing information about weapon design. Enclosure 3 defines the justification for withholding this information, as well as instructions for submitting an appeal of the DOE determination.
Please be advised that one document originated with another agency. Because we are unable to make determinations as to the releasability of other agencies' information, the document has been referred to the appropriate agency for review and direct response to you.
If you have any questions regarding the denial of ODNI information, email our Requester Service Center at DNI-FOIA@dni.gov or call us at (703) 874-8500.
Sincerely,
-kv f~trifer Hudson
{)
if:)i;ector, Information Management Division
## Enclosures
# Sharp.2009
Transforming Nuclear Attril:w.t'f~;J
Culture, Community, ;.3nd;;t~:paljge
## Sec~;
| (U) | EXECUTIVE | SUMMARY | 5 |
|------------------------------------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------------------------------------------------------|----------------------------------------------------------------------------------------------|-------------------------------------------------------------------------------------------|
| (U) | | | |
| EXECUTIVE | SUMMARY | _, | - 5 |
| (U) | KEY | FINDINGS cceecrereireerarssnsnssansmssnrassasrsnsessnesssstasasassan | ssssrsessnsnnssassssessnssesssassesnsrsnsessssnast |
| (U) | Developing | the | Culture eeecnen... |
| (U) | Developing | and | Leveraging |
| (U) | Developing | the | INfrastriiClure ecoevmenierrcraereseessesecrsterensnssssnesisassessssansansesssessaes |
| (U) | Developing | the[____ | tructure |
| (U) | | | |
| KEY | | | |
| FINDINGS ...... ........ ........ ........ | | | |
| 6 | | | |
| (U) | | | |
| Developing the | | | |
| Culture | | | |
| 6 | | | |
| (U) | | | |
| Developing | and | Leveraging | |
| People 7 | | | |
| (U) | | | |
| Developing the Infrastructure | | | |
| 7 | | | |
| (U) | | | |
| Developing thec==:ptructure | | | |
| | | | |
| 8 | | | |
| (U) | SCOPE | NOTE | 9 |
| (U) | | | |
| SCOPE | NOTE | - | _., |
| (U) | INTRODUCTION | TO | THE |
| (U) | | | |
| INTRODUCTION | TO | mE | ARTICLES |
| (U) | ASSUMPTIONS | 12 | |
| (U) | | | |
| ASSUMPTIONS | _.,, | - | |
| ~ | | | |
| ... | | | |
| 11 | | | |
| (U) | SPECTRUM | OF | THE |
| (V) | | | |
| SPECTRUM | OF | THE | THREAT |
| - ,_ | | | |
| ~-,."---13 | | | |
| . | * | | |
| ~ | | | |
| (U) | NEED | TO | COMMUNICATE |
| (U) | | | |
| NEED | TO | COMMUNICATE | - |
| i~~~- | | | |
| | - | .. | |
| 15 | | | |
| !'!' | | | |
| ':!i.1o~~ | | | |
| (U) | COMMUNICATION | CONCEPTS | coveeeerrencememscscnscrnersserans |
| (U) | | | |
| COMMUNICATION CONCEPTS ....... - .... ..... - ........ | | | |
| ~~:~::~15 | | | |
| (U) | | | |
| COMMUNICATING A TIRIBUTION ASSESSMENTS . . . | | | |
| ~~~l~li~:~1~!~!i! | | | |
| | | | |
| I | | | |
| 5 | | | |
| (U) | COMMUNICATING | ATTRIBUTION | ASSESSMENTS |
| (l!fr))rm.Lna1.1ch.m.msf..um..."..n _.mI | | | |
| ( | 1 | | |
| o) | ; | S | |
| ~"'B!JC- | r::::::::;::::::::~~::::~:::::~;i::;1f~:::::::: | | |
| i! | | | |
| | . | | |
| ~ | | | |
| .. | | | |
| ~l::;li~-J. | ~;r | | |
| (U) | | | |
| EXPRESSING UNCERTAINTIES IN | CRISES | - | |
| !i:r.~ | | | |
| | | | |
| ~~i~t:L | | | |
| | | | |
| 19 | | | |
| (U) | EXPRESSING | UNCERTAINTIES | IN |
| B, | K | | |
| (U) | A | PROPOSAL | FOR |
| (U) | LESSONS | FROM | PREVIOUS |
| (U) | | | |
| A PROPOSAL FOR DEVELOPING STANDARDS | OF | UNCERTAINTY | |
| ~i(;-i-HE'D.: | | | |
| ... | | | |
| :t | | | |
| 20 | | | |
| (U) | | | |
| LESSONS FROM PREVIOUS STUDIES ON UNCERTAINTY | .. | ,:: - | |
| .'t | | | |
| ::~t&!f: | | | |
| | | | |
| 23 | | | |
| .::~~~~:;~ | ~~- | | |
| (U) | ORGANICALLY | GROW | A |
| (U) | | | |
| ORGANICALLY | GROW | A LEXICON | |
| ~~Ji*1~ | | | |
| .. | .::.. | | |
| ,~,~~ | | | |
| | ...:.. | | |
| 15 | | | |
| (U) | SPLIT | SECOND | DECISION-MAKING: |
| (U) | MITIGATING | THE | "CS1 |
| (U) | LE/IC: | RisK | OF |
| (U) | | | |
| SPLIT | SECOND DECISION-MAKING: | | |
| A:if!~w | E~;f.:ORCE;f(l!f,tf.ERSPECTIVE | | |
| .. | ---- | 27 | |
| . | " | | |
| ~~:::~~;~~- | J~; | | |
| . | | | |
| ;t | . | | |
| (U) | | | |
| MITIGATING THE | "CSI | EFFECT | |
| . | | | |
| ~ | | | |
| - | | | |
| ~~::;i,,~ | | | |
| ... | | | |
| ,~:~: | | | |
| | | | |
| 21 | | | |
| (U) | | | |
| LEIIC: | RISK | OF BEING | WRONG |
| :!'J,ti{i~;: | | | |
| 28 | | | |
| (U) | MINEFIELDS | AND | MIND |
| (U) | | | |
| MINEFIELDS | AND MIND | TRAPS | IN |
| ~- | ~it!~:~ | | |
| (U) | THE | ROAD | TO |
| (U) | | | |
| THE | ROAD | TO | HELL |
| WIT~J~B~tiVE | | | |
| COGNITIONS | .. - | 31 | |
| (U) | ACCOUNTING FOR | | |
| PSYCHOI.C!9.~~:f'-ND:~IAL | | | |
| FACTORS THAT DEGRADE RESPONSE | OR | INHIBIT | |
| INFORMATION SHARING | | | |
| IN | LJANAJ:YSIS | ...... ........ ........ | |
| 31 | | | |
| (U) | | | |
| Core Concepts in | Judgiijent,.,a'nd | | |
| D~ion-Malcing | | | |
| | | | |
| 31 | | | |
| ~~Ji.~~:::: | | | |
| .. | | | |
| ~::::::::.:::: | | | |
| .. | | | |
| :.:~.:::::::: | | | |
| | | | |
| ::::.::::.:.:::::!~ | | | |
| (U),:Hnplicit.vs: | | | |
| Explic1t~i}ion-Making | | | |
| . | | | |
| | | | |
| 35 | | | |
| ... | | | |
| ,r!~~-~ | | | |
| . | | | |
| ~ | | | |
| .. | | | |
| ~ | | | |
| ... | | | |
| @}.The Haiiihis | of | | |
| Bias.~lfeuristics, | | | |
| and | | | |
| Intuition 35 | | | |
| | | | |
| ~... | | | |
| ,. | | | |
| ? | | | |
| . | " | | |
| M:IJ.FJ.ie~:Yfre | | | |
| Possessions' | 36 | | |
| (U) | | | |
| ';4ifc.~~ring | | | |
| and | | | |
| the Pressure to Round up the Usual Suspects | 36 | | |
| ~ | Z~7~~~:~:.~~~::::::::i~ | | |
| (U) | ACCOUNTING | FOR | PSYCHOLOGICAL,AND; |
| INFORMATION | SHARING | IN | I___:I'ANA;,YSIS |
| (U) | Core | Concepts | in |
| (U) | Rationality: | Bounded | and |
| (U) | Biases | and | Heuri$fics |
| U | Baunded | Wzllpower | TR |
| (U) | Boun | EESL | ..v.eeteesesemsessaeseesessnesssbuse |
| (U): | 'Imphcmvs | Exphc'itfl?egz,flonMakmg | |
| (l'{) | The | Hazara's | of |
| - | (U)r'Bg!zefi | ATE | POSSESSIONS |
| ) | Am:honng | and | the |
| ) | How_,sWe | PErCeiVe | RiSK cccovvureemmsiineinisniinnssrer |
| (U) | Hot | Cogggnon | |
| (U) | Stress | Efi'ects | on |
| (U) Stress | | | |
| Effects | | | |
| on Analysts | | | |
| and | | | |
| Policy-Malcers | | | |
| | | | |
| 38 | | | |
| , | | | |
| (U) | PERILS | AND | PITFALLS |
| (U) | | | |
| PERILS | AND PITFALLS | OF | GROUPS - _ _ |
| (U) | | | |
| SUCCESSFUL CULTURE IN | HIGH | STRESS, HIGH AMBIGUITY ENVIRONMENTS ... .... . | 41 |
| (U) | | | |
| Recommendationfor~re | | | |
| - | 41 | | |
| (U) | | | |
| SOCIAL NETWORKS AND | | | |
| TilE | | | |
| 42 | | | |
| (U) | | | |
| GROUP | DYNAMICS AND | MA | ONS .. - ... - . ... ........ |
| (U) | | | |
| "Groupthink" | | | |
| | | | |
| 42. | | | |
| (U) Stage | of | Development 43 | |
| (U) | | | |
| MEMBERSHIP | - - - .. | ; | .... .. ..... |
| (U) | | | |
| Multi-disciplinary | | | |
| Advantage | | | |
| 43 | | | |
| (U) | SuccesSFUL | CULTURE | IN |
| (U) | Recommendation | for | ulture |
| (U) | SOCIAL | NETWORKS | AND |
| (U) | GROUP | DYNAMICS | AND |
| (U} | "Groupthink" | | vereeerser |
| (U) | Stage | of | Development |
| {U) | MEMBERSHIP... | reeeerecvernanaesesees | : |
| ) | Mulhdzsctplmmy | Advanlage | e |
| (U) | EXPERTISE: THE | GOOD, | THE |
| (U) | The | All-Star | Problem. |
| (U) | All | Expertise | is |
| (U) | EXPERTISE: | THE | GooD, |
| (U) The All-Star | | | |
| Problem | | | |
| 44 | | | |
| . | | | |
| (U) | | | |
| All | Expertise is not the | | |
| same., | | | |
| 45 | | | |
(U) WHAL iS EXPEIrISE? .......ooeeeanmeeeesiressraeasesersessassssamenssnsssstsstmsnssnssmsstastensmmmsessessssasnsanenn
(U) Self-herding Cais?
........ouceirireecncenecsnssenessossssmonsasesssesssisasssessassssnssassssesssissens
(U) ROLE OF NEGOTIATION ......c.coooenenmecesrerssssnseseemmmaseesesesesemsesessssemssssesesmmmsseesessessensesssersssmsesessesssermnossers
~ (U) THE CRITICAL ROLE OF EXERCISES AND TEAM-BUILDING ........
(U) Recognize the Need for
Inoculation of Personnel and Information Networks Prior to Evenis....
(U)Establish Metrics to Ensure Effectiveness of Exercises and Training.......
-
(U)Preserve Splits that Present Differences of Opinion to Policy-M
(U) COUNTER DEBILITATING EFFECTS OF HIGH STRESS ENVIRONMENT
MONITORS
(U) Emphasize Informatmn Sharing and Enhanced D Coardmanon ----------------------------------------------------
(U) What is Expertise? ........ 45
(U) Self-herding Cats? ........ 46
(U) ROLE OF NEGOTIATION ........ 47
(U) THE CRITICAL ROLE OF EXERCISES AND TEAM-BUILDING ........ 47
(U) Recognize the Need for 0Inoculation of Personnel and Information Networks Prior to Events .... 47
(U)Establish Metrics to Ensure Effectiveness of Exercises and Training ........ 48
(U)Preserve. Splits that Present Differences of *Opinion to* Policy-M~
........ 48
(U) COUNTER DEBILITATING EFFECTS OF HIGH STRESS ENVIRONMENT L___j BY EMPLOYING STRESS
MONITORS ........ 49
(U) Emphasize Information Sharing and *Enhanced* 0
Coordination ........ 49
(U) BRINGING THE RIGHT BROOMSTICK TO THE WIZARD
(U) BRINGING THE RIGHT BROOMSTICK TO THE WIZARD ........ -------51
(U) UNDERSTANDING THE SENSITIVITIES AND CONSTRAINTS OF SENIOR POLICY-MAKERS
(U) The Efilzz{wt as Consultant ........cuenevceneneenn,
(U) The
t Ask the Right Questions ........cousveene
------------
(U) UNDERSTANDING THE SENSITIVITIES AND CoNSTRAINTS OF SENIOR POLICY -MAKERS ........ 51
(U) The~yst as Cons~ltant ........ , ........ ; .... 51
~
The
tAsk the Ri ht
lions ........ ::::::::~:::::::j~{~:::::::: ~~
L
(U) Recognize t t
essure or Imm iate
rt- erm
o ltical or *Policy* Responses'Will;:P,~o.:Pire,f!tly
Proportional to the Magnitude of the Consequences Surroundmg an Event
22
(U) Recognize that Policy-Maker Tolerance of Risk (and willingness to accept varioiis;measures or
estimates of certainty) Will Diminish as the Magnitude of Potermal
Conseqqences (On Extl:er Action or
Proportional to the Magnitude of *the Consequences Surrounding an Event ........* ,,Nl:"""'"~1~it:~,;F .... 55
(U) Recognize that Policy-Maker Tolerance of *Risk (and willingness to accept* iznO~;/!!f!OSUreT-6r
estimates of certainty) Will Diminish as *the Magnitude* of Potentia~ Conseq'f~'es (O,tl~!.i!l!lf...Action or
Inaction) Increases ........ f.,'!Jt: *........* :.i',-:~:~~:.; *........* :~~: ........ 55
(U) Not All Senior Policy-Makers Are the Same ........ > *........* .'~:;:;~: ........ ::: ........ 57
n 11 Th Co
*A th
b'
~;!.
-~~tt~!:~ .. l!.-..,-
9
I'-'./
e
ntract1on OJ. u only *pro* .em ........ ~.':+;;z.: ........ :~ . ;::: ........ 5
(U) Using the Right Kind of Expert Advisers to *Interoct With* P~f.if;fmaliefo~~ ........ /:'. ........ 60
"d'i't
Inaction) Increases........
i
(U) Not All Senior Policy-Makers Are the Same...
(U) The Contraction of Authority problem...
) Usmg the Right Kind of Expert Advisers to Interact With Pah;:y-make
:
The Role of Blame Avoidance Strategies and How These May Comphcate,tke Atmbutzon Questxan Jor
........
(U) "Safety in Numbers?" The Hurricane Katrin& Rg;oblem i (U) The Surrounding Psychological Context for:bnd Pohcy-Maker in WMD Attribution........
(U) Terrorists are "Adaptive Enemies" Who w;ll try t ) "Muddy Ihe,'Waters " to Complicate the Attribution
Process by Increasing Ambiguity
w~-~~~~-~!.~~~~~.:7-~~~-~':;~~~e:...~~~-:::=~~~:~~~~~~~~~~~~~~~~~~~~.~~~~~~.~~
(U) 'Sqfety m Numbers.
The Hurncane Katrif!_lt;'IJt.iJ.~{em ...... ,:f::;:; ........ 62
(U) The *Surrounding Psychological* Context.fo~
Policj.~'in WMD Attribution ........ 62
(U) Terrorists are '"Adaptive Enemies" Wlfp;w;tl *try* trf.~ Muddy the;Waters" to Complicate the Attribution
Process by *Increasing* .Ambiguity ........ :(::~:~~!:~i.:~~):~ ........ : ........ 63
(U) THE RIGHT STUFF: HARVESTING EXPER'f.i~.1~ ........ --........ __, __ , ........ --........ 65
(U) THE RIGHT STUFF: HARVESTING EXPER'I' SE
2, l
E
........
(U) INTEGRATING EXPERTISE
'
.......
(U) ANALYTIC ROTATIONS FOR CULTURAL AWARENESS X
------------
(U) INTEGRA TINO EXPERTISE ........ :~;;f;!t:< ........ , ........ 65
(U) ANALYTIC ROTATIONS FOR CULTURAL AW.~ESS,.;~: ........ 65
(U) DRA WINO ON THE "UN-cLEARED" Wo~::: ........ 68
(U) DRAWING ON THE "UN-CLEARED" WORH_I
------------
(U) RED TEAMS ARE Essmmm i,
------------
(U) EXERCISES AND TRAINING ..., 0.
(U) COLLECTING AND PRESER\;ING?ERISHABLE INFORMATION ........
~~~~~::s~~~=.:i=~~~::;:1~~~:~:::~~::::::::~:
(U) COLLECTING AND ~ER~P'PERISHABLE INFORMATION ........ 70
":fo;lo
.:!.;;h
=t~
(U) INFORMATION SHA'RING%MONG DlSPARATE KINGDOMS
.'.,.-
(U) INFORMATION SJI#fUNGl~~1'.~~msp ARA TE KINGDOMS .. --.. -........ ---... - ........ 71
------
(U) THRO}\!JNQ,JliEJii!.:S.~H .... ~.~~::~.:. ........ : ........ 71
(U) C'.'~~Rni9its>w C6~~~~SE DATA AND AssESSMENTS ........ 72
------
W) CAFI'URING Low CONFIDENQE DATA AND ASSESSMENTS
(9)) CQNT]NUITYOF CONNECI'[.VITY
(U)-L-,EvERAGnJG THE CATALYST PROJECT
(W) lT{N'rEnEACEs THAT ENHANCE INTERACTION
..
o
---------
()] DATA: SHARING VIA COMMON DATABASES........ccccovnisivrnens
) lNFORMA'flQN SHARING THROUGH SOCIAL NETWORKING TOOLS
Sh:1t.;~;ii~~~~R~~~~:::::::: ;;
(U) IT~ACES THAT ENHANCE INTERACTION ........ 73
(U) 0ATA~~~NO VIA COMMON DATABASES ........ 73
(U) INFORMATI9}:1~SHARINO THROUGH SOCIAL NETWORKING TOOLS ........ 73
!~;;~--
(U) THE MIXED 'REALITY PERSPECTIVE
(U) THE M~p'REALITY PERSPECTIVE.-........ - ........ - ........ _ ........ 75
(U) INTELLIGENT AGENTS.....crovermeurssssasnsnsnsnes
------------
(U) OVERCOMING CLASSIFICATION BARRIERS
(U) ENHANCING DATA MINING
........
........ ........
(U) DYNAMIC ANALYSIS PROCESS FOR THE[
Il
-----
(U) INTELLIGENT AGENTS ........ 76
(U) OVERCOMING CLASSIFICATION BARRIERS ........ 76 (U) ENHANCING DATA MINING ........ 77
(U) DYNAMIC ANALYSIS PROCESS FOR THE o ........ 77
(U) LEVERAGING LOCAL ASSETS
(U) LEVERAGING LOCAL ASSETS ...... -
........ ~ ........ ~ ...... - ....... _ ........ __ ........ 79
(U) INCREASING EFFICIENCY
.....ccostivenrunensaesnenerersnsssassesessssnsasnssssarens
(U) TEAMING WITH LOCAL LAW ENFORCEMENT
(U) INCREASING EFFICIENCY ........ : .... 79
(U) TEAMING WITH LOCAL LAW ENFORCEMENT ........ 79
(U) CAN OTHERS UNDERSTAND YOUR DATA?
(U) CAN OTHERS UNDERSTAND VOUR DATA?-........ _ ........ 81
(U) MANAGING THE DATA
....vevreerneceeneremenassasssnnes
(U) DATA CHARACTERIZATION AND ANALYSIS ........
(U) MANAGING THE DATA ........ 8 I
(U). DATA CHARACTERIZATION AND ANALYSIS ........ 81
BY EMPLOYING STRESS
........
........
AR AN
#'.ln.
........
--------------------------
63
--------
........
(U) DATA MODELING AND VISUALIZATION...
Lessssmsarescsssarsres 82
(U) USING METADATA TO FOSTER INFORMATION SHARING AMONG SEPARATE COMMUNITIES ........ovvereenn.. 83
(U) DATA MODELING AND VISUALIZATION ........ , .... : ........ 82
(U) USING METADATA TO FOSTER INFORMATION SHARING AMONG SEPARATE COMMUNITIES ........ 83
(U) HOW TO BEST USE TNF INFORMATION
85
(U) HOW TO BEST USE TNF INFORMATION ........ m._ ...... .-........ -------85
(U) MANAGING EXPECTATIONS: WHAT CAN WE EXPECT FROM TNF AND WHEN CAN WE EXPECT IT?........ 85
(U) Pre-Detonation vs. Post-Detonation
terereretnesareerfesnestranesarabestearereeaeest
s be s enanebasastensentans
.86
(U) TNF Support 10 IND AUTTOULION. ........ceueeereeeverriserivenssnerisnsnssessassssvsrssssassssassasassesnsassssasasasnss 86
(U) TNF: Common MiSconceptions........ewveevuiuisnsesens
esesecerasre
e s
e
s ae s e sastnresanesrere 86
(U) TNF: Exercising t0 DiSpel the MYIRS ........c.ocioioimieirerreeressssisiscsssasseeiasnssassessnsnsessssassassmssssansn 87
(U) LIMITING BIAS IN TECHNICAL DATA INTERPRETATIONS
.
-
89
)
haring Information with the TNF Team........iceeoeectecinicceseeeeaeesassssnens 89
(U) Incorporating TNF Results into [
|ASSESSMENLS ........c..coviimirisiinisssnssisisssinsssessissssnssss 90
(U) MANAGING EXPECTATIONS: WHAT CAN WE ExPECT FROM TNF AND WHEN CAN WE EXPECT IT? ........ 85
(U) Pre-Detonation vs. Post-Detonation ........ 86
(U) TNF Support to IND Attribution ........ 86 (U) TNF: Common Misconceptions ........ 86
(U) *TNF: Exercising to Dispel the Myths* ........ 87
(U) LIMITING BIAS IN TECHNICAL DATA INTERPRETATIONS ........ 89
(U)l
~haring Information with the TNF Team ........ 89
(U) Incorporating TNF Results into *r=JAssessments* ........ 90
## M) Putting It All Together: Proposed Function And Structurh I
~~-~~~~~~~-~~~~~~~:.~.~~-~~~~-~~~:~~~~-~~-~~~~~-~~::~7--91
| | | (U) |
|----------------------------------------------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------------------|-----------------------------------------|
| WHEN | TO INITIATE THE FORMAL PROCESS | |
| ;~::.:.:;::, | | |
| | 92 | |
| ~~~ROPOSEoc:J<FsPONSIBILIIIES | | |
| | | |
| J::::::::;~~~;~~;~:::"::~~~~~~;;J:::::i | | |
| (U) | | |
| PROPOSED STRUCTURE OF | THEC] | |
| ! | .. | ;: |
| 93 | | |
| (U) | | |
| A | Two- | TIER STRUCTURE |
| ~~~:: | | |
| | | |
| :!;~~~~;:;:.: | | |
| | | |
| 93 | | |
| (IJ)f~!:'::::::.::::::::l::::::::~;~~~~~;;::i~~::::::::E | | |
| 1Y)' | | |
| ) | | |
| (U) | PROPOSED | STRUCTURE |
| (U) | A | TWO-TIER |
| (U) | " | IExecutive |
| (U1 | Support | Group u...onneneceiieveeeesressivirenen |
| W | | | ] |
| (U) | REQUIRED | ACCESS |
| (U) | | |
| REQUIRED ACCESS TO INFORMATION ... ........ | | |
| ,[:-~ | | |
| . | : .... | |
| ::;,;,;,;~ | | |
| | | |
| 97 | | |
| (U) | BUILDING | THE |
| :'"" | | |
| ";!~:':i:i-;:0.. | | |
| . | | |
| (U) | | |
| BUILDING | TJIE | NUCLEAR ATTRIBUTION |
| -.:~;t: | | |
| | .- | |
| 100 | | |
| (U) | TRADE | SHOWS: |
| {U) | BRIDGING | THE |
| (U) | COMMUNICATE | TODAY; |
| (U) | WHEN | REALITY |
| (U) | | |
| COMPETITION | | |
| 1S | | |
| GOOD... | | |
| (U) | OuTLOOK | . |
| (U) | | |
| TRADE SHOWS: VIRTUAL AND REAL . | | |
| ;.;:;:;;;;~~~~=~~r!:::::::~;;~t;;.- | | |
| ... | | |
| ::~ | | |
| | | |
| 100 | | |
| (U) | | |
| BRIDGING THE NUCLEAR EXPERTISE | | |
| GENE~TION | GA"f:~:~::t~;~~f';: | |
| ... ....... | | |
| 101 | | |
| (U) | | |
| COMMUNICATE TODAY; STILL | | |
| CoMMU!'il~~j~!J | | |
| TOM.gRROW :;':-: .... .. ........ | 102 | |
| (U) | | |
| WHEN REALITY COSTS TOO MUCH, | | |
| Go"VnitJj_;\1,.,.,,;1:~ | | |
| .... - .. - ... | 102 | |
| ~ | g=:;~-~~~-~~::::::::~~~~~~~t~~;~;;:::::::: | :~~ |
| - | .. | |
| !._.~~ | | |
| (U) | SUGGESTIONS: | REQUIRED |
| e | | |
| (U) | | |
| SUGGESTIONS: REQUIRED | | |
| RESOURC~$, | | |
| :::; | ,_ | - .. 104 |
| (U) | FUNDING | NEUTRAL? |
| (U) | GETTING | BLOOD |
| (U) | | |
| FUNDING NEUTRAL? | | |
| ;;;:;:,!~;~~~ | | |
| | ::. | |
| 104 | | |
| (U) | | |
| GETTING BLOOD | FROM | |
| A | | |
| STQNE~i;:.~:: | | |
| .. | :-:: | |
| (_U) | | |
| SCRUBBING FOR RESoUR<:;Es | ... | |
| ,.;:~:.- | | |
| | :, | |
| (~F) | | |
| AIRCRAFT | ANDAFTACSi:iPPORT.;: | . |
| UJ) | SCRUBBING | FOR |
| (SHINF) | AIRCRAFT | AND |
| (U) | How | MANY |
| (U) | | |
| How | | |
| MANY~~~~~~~:~~~?:: | | |
| | . | |
| (U) | APPENDIX | A: |
| (U) | | |
| APPENDIX A: | | |
| INQ~S. | | |
| OF | RECOMMENDATIONS AND | FINDINGS-- |
| | | |
| ~.,.r | | |
| ... | eoii. | |
| 0]:.-;.~i~l!! | | |
| I | -A"'r'l'mgy;rlon | TERMS |
## L8S) Appendix C Dynamic Analysis Process
(U) APP~IxipJ;;ATm;!{fiQN TERMS ........ _ ........ -
........ 117
-
.~;~&i
.
~:~~~
~~: t:,~:o.. ...
(U) "-P.P,Ji;NDIX (J:'DYNAMl'ANALYSIS PROCESS ........ _ ........ - ........ 138
.~ ... ~-~":w!!-:
'.
.
(U) APPENDIX D. SHARP 2009 NUCLEAR ATTRIBUTION PARTICIPANTS AND PERSONNEL. 141
(u). APPiOO?,-.x'D: SHARP 2009 NUCLEAR ATTRIBUTION PARTICIPANTS AND PERSONNEL.I41
(U) POLICY~fiocuMENTs ........ - ........ _ ........ - ........ 144
::!!';;,;,~ .... ~
-~:!!i!.;!;:"'
(U) POLICY bQCUMENTs
144
(U)END N(mafi"f""
144
(U) END NOTES:::. ........ _ .. , .... - ..... --.. - ........ _ .. _ ........ ---144
..
,.
## (U) Executive Summary
(U) A group of 30 experts from the government and private sector met in Tempe, Arizona in August 2009 to study the topic of nuclear attribution under the auspices of the Summer Hard Problem Program (SHARP), sponsored by the Office of the Director of National Intelligence
(ODNI). Participants included intelligence analysts, members of law enforcement, scientists, academics, and subject matter experts in national security policy, proliferation, terrorism, law, crime, behavioral psychology, and other specialties. (See Appendix D for a list of participants). The study focused on communication challenges confronting three distinct communities involved in preparing nuclear attribution assessments for executive branch leadership. Information coll~ted and analyzed by those communities -
technj~cat,nuclear forensics (TNF), law enforcement (LE) and Non-Title 50 organizations, and tti~~Jt.ltelligence Community (IC) - must be fused to formulate an overall attribution assessment'-"~1";-time.-
crisis environment with mostly incomplete and evolving information.
.,;.:.:...
'!::;~i:.~;;
_.;:;~~:!~~~~.'
~;'"~~
(FOoo)At SHARP, the participants role-played seve~l n~1ear a~J?;4~ion stEit)fibs while immersed in microcosms that combined law enforcement, ffi"telligence;;:i:nd techilical communities. Leveraging this. microcosm environment, the p~~,~pantSw;~lasked with creating their best all-source attribution assessments. The~:;PcFi:~ces of identifying, developing, working with best practices for attribution W: ,,tpe SHARP;r~fcrocosm enabled the participants to scale up their findings ~!~~~iy]i~_& success'of deploying a new national attribution capability (e.g. an L_ .. ~_f'Til~~:!JARP;'~~Y..(mdings fall into four broad areas: culture, people, information sha$.t!jpfrastnj.Cture, andLJfunction and structure:
.! *...* , .. ._,':'~~.;.~
.. r..'
'
- . Culture: At least three distinct conimuillt,i~.oi'"kingdoms" (IC, LE, and TNF) will be
involved in a nuclear attribution investigl:iii~,9;.,;, SHARP recommends several measures,
such as information sharing, team-building, jq:~t.exercises, common training, common
lexicons, and persistent social networlc:ijlg. Sl;!Ch measures will institutionalize the right
practices, behavioral norms, .!!!!d cgUilboration that will be required of a multidisciplinary
team working seamlessly:.~=jj{OOuc~ attribution assessments.
,..
....
. ..
-
People: SHARP sugges~ tJ:uit. a pr~gram of training, rotational assignments, and
mentoring be deveJ,9:ped:~;~F2,~~il@ build a sustainable cadre of cleared analysts with the
right expertise tQ:\V,ork the:nuclear attribution account. Additional cleared specialized
expert,!should.o~;-l~eraged through a virtual architecture, and expertise from the open
so~c~";6ti,~;~hou1a:[~~J~veraged through technical means such as crowd-sourcing.
-
.. IMdrmatiori;:Sharing Iiifrastructure: SHARP recommends that a formal study be
. cond~~~'io more comprehensively identify and prioritize options for mitigating
infoniill~9P. shapng impediments. The normal way of doing business is too slow and
exclusive:t9!(1}iing the full capacity of the distributed law enforcement, intelligence, and
technical ~hununities to bear on the problem in the quickest manner. Tools and
approaches must be developed and deployed now that, when activated for a nuclear
emergency, allow relevant players to share knowledge at the speed of technology, not at
the speed of bureaucracy.
-
unction and Structure: The SHARP study developed a two-tiered model of the
consisting of an executive level and a support group - and recommends adding
two new responsibilities for the c=J I) Develop recommendations for communicating
crisis-related information With government entities (Congress, state, local, and foreign)
and the public. 2) Establish a means to interface with consequence management
constituencies.
(U) The findings of this SHARP are applicable to any operation where disparate communities
~
or teams must work together to be successful in solving difficult and complex tasks.
## (U) Key Findings (U) Developing And Leveraging People (U) Scope Note
(U) This report presents the findings and recommendations from the SHARP 2009 monthlong session, "Assigning Attribution in Nuclear Forensics and Intelligence Analysis." The nuclear attribution SHARP session engaged approximately 24 total government and external experts, including nuclear scientists, forensics examiners, policy experts, cognitive I
neuroscientists, behavioral scientists, proliferation experts, and intelligence and law enforcement analysts. The SHARP report represents the views and analytic fmdings of a diverse group of participants in tackling the hard problem of how to integrate three information streams - technical nuclear forensics (TNF), law enforcement (LE), and the intelligence community (I C) - to formulate rapid and credible assessments, ~4icl}.;may be based on incomplete or time-evolving information, in support of nuclear attribiltfbn.
,!'!:"':!::!:" ...
~
~:~~~:~;,~
i~
(U) The study focused on identifying strategies and solutions to overco~~i~~~lenge&;:;.n:
producing effective multi-community-authored assessments suppo~:t].g;attri6iij~Qn. ,LE, TNF, and IC each use distinct terminology and has its own way 9.;~~hinki~~~~put the"'P.t6blem and manner of assessing uncertainties. Yet, their assessments and wpclusiO'iiS.P.a~e"io be fused and presented to policy-makers cohesively and cogently in ~!9ef;!~.;.
~~~~~v
. ;~:::~
~~~!f;~~;t:~ ~
-
Provide the maximum amount of info~~~~n ~~j~ight as t<i'the source and
perpetrator( s )
'~ ''-li!,;
. ,.,,,._
'
... :;~~
. "{~::!~
,:~~:~.;ov'
-
Convey the results in a form that.ti~~t meetS:The need ofleadership;
-
Make as clear as possible the c~~fiii~~~l~~els and w;~ertainties that should be
attached to findings and
:~l:~;:;,
'
-
Effectively present alternative analyses ~~(explanations for available information.
(U) The final report is structured to convey fmdings and recommendations to the DNI on key factors that play in effective integration and communication of multi-community attribution assessments. As a result, the fmal report is presented as a compilation of articles, each addressing a key factor in the attributio~~~-~e SHARP participants self-assembled themselves into teams to write the articles within, and in many instances, participants contributed to several articles.
(U) The findings from the SHARP 2009 Nuclear Attribution session are applicable to any mission where disparate communities must work together using incomplete and evolving information, and the conclusions must be effectively conveyed to decision-makers.
## Se~ (U) Introduction To The Articles
(U) The SHARP report is a compilation of articles that address key issues associated with successfully conducting multi-community-authored attribution assessments that would then be communicated to decision-makers. Taken as a whole, the articles would assist in developing and deploying a seamless team to producing rapid and credible attribution assessments based primarily on the information streams from the technical forensics, law enforcement, and intelligence communities.
(U) ASSUMPTIONS
## (U) Expressing Uncertainties In Crises (U//F<Th..O)Ir----------------------.1
L__ ________
__.jl Specifically, the intelligence, LE, TNF, and P.Q\ifY cotilJP.l!l:rl'ties
have each, over time, established different lexicons utilizing quantitati~~:ruidif,tqalitah~e
descriptors to express uncertainties. Dr. Charles Weiss has;~ategoJ#~~ ~orne 6f:~ese
disparate lexicons and summarized them the table below1, ~::
-' :::::.~
/
.
~=~=~:~
;::~~=,-~
Given the iif~~~9f..!lcommon objective standard for expressing uncertainty, policy- and decision-makef;"often are unsure how to interpret the meaning of information being conveyed to tliem or what degree of action the analysis warrants.
(U) The discussion of uncertainty usually centers on two distinct methods of expression. One is scientific uncertainty based on statistical analysis of numeric data, utilizing measures such as mean, median, standard deviation, and confidence interval. The second is subjective uncertainty, which does not lend itself to quantitative expression. In most cases, policy- and decision-makers are dealing with subjective expressions of uncertainty. This presents a difficulty, since subjectivity results in inconsistenc~es in interpretation.
(U) For nuclear-related events, significant uncertainty may persist for some time, as the sCientific/statistical findings may not be available early on and will be subject to change as
## Sec~Orn
the data are analyzed and interpreted. TNF findings are a combination of comparative statistics from isotopic ratios, measured levels of trace elements/molecules, shielding and packaging materials, and ancillary artifacts associated with the material package. The comparative isotopic ratios are based on analyzing the degree of agreement between measured ratios for the sample in hand and databases on such ratios for worldwide nuclear materials. Experienced nuclear scientists then make judgments on the most likely associations, taking into account other artifacts of the sample.
(U) The judgments resulting from this process are presented in terms ofthe most likely to less likely matches to various possible sources, including some indication of confidence level.
The confidence levels are a combination of the statistics of measurements and seasoned judgments on bow closely the data fit with the sources in the worldwide infon.nati9n base.
Conveying this process as background to a decision-maker in a crisis may not;~~: practical, but I
I should be prepared to do so in various levels of detail.
,:.-:;\::.,. _;
(U) Intelligence analysis is often a qualitative process involving the int~~ii~p._.9f o6~~~ed facts, the views and opinions of government officials, and w.fonnati~q,jrom llii'fiiilJl-'sources of varying reliability, in the context of historical data and c'lifumon kiiciwl~ge .. Rarely does a
.
-
""'If'
single observed fact drive a conclusion. Each of these sources. of.inforniatibn'has an associated level of uncertainty.
. .:.:::; /'.;:~;;:;~~.
_:::
.,1! .. :.:-::.i"
(U) Law enforcement combines laboratory ev\4~~!! 'ifij;~~issible;~i~cumstantial evidence, eyewitness testimony, information from infofiiiij;:and d.~limentary evidence to build a case against a defendant Measures oftW~~,rtainty:for the Ieg~i~ystem are referred to as
"standards of proof." In the criminal justi~:~Y.S~~iii. convictio"n requires the highest level of certainty: proof beyond a reasonable doubt.;.;~;;::;,.
'
:"!::-;-..
..
(U) A challenge for the integrated IC, LE,.Mtd ~~~mmunities, then, is how to aggregate the various means for reaching conclus_j5?_nj' with {meaningful way to present levels of confidence. A common lexicQ~::(qr,.expressing uncertainty and confidence would be ideal.
However, an externally imJ)9~ecJ:~hol~sale transition to a new, common method of uncertainty expression !~-~i.kely to ~l)cceed.
as a newly-established entity, has an opportunity to es~~Jish a;~~~~finethod for the expression of uncertainty. Approaches to doing thi~ ~!! de_t~.i! in the section entitled (U) ORGANICALLY GROW A LEXICON .
.
~::,.:::~~:!::. . :..;::'"' .. ~
(U) 'Ift~:~~~~-i~~fnent on~pPtlunOn lexicon will ease the work I
I and enhan~e comj)R:Jl'ensionby decision-makers. It also provides a basis for a common language of uncertafu'ij;:t'fross all communities. This will not require that the individual communities abandon tl:ic;&:current standard approaches, but will facilitate enhanced collaboration .
. ~-:;~:::~
(U) A Propos; for Developing Stant:lanls of *Uncertainty\L* ---=--=----'
(U) Of the approaches to stating uncertainty, the legal standards might provide a starting point for expressing aggregate uncertainty in Nuclear Attribution. These legal standards or equivalents could be used to represent confidence associated with the combined inputs from LE, IC and TNF. The legal model also may be useful to the policy community in terms of standards of proof that should be met in recommending actions that employ national and domestic security levers of influence. These levers include, for example: judicial, diplomatic, intelligence, military, and economic.
(U) The US judicial system uses a time-tested way of conveying uncertainty in reaching judgments on criminal and civil cases. These uncertainties are represented by standards of proof such as beyond a reasonable doubt, clear and convincing evidence, and preponderance of the evidence. Furthermore, standards of proof directly relate to verdicts which, in turn, a judge uses to decide on sentencing in criminal trials. The jury is made aware of what level of uncertainty must prevail in rendering a verdict on the basis of all evidence, testimony, and arguments presented during trial proceedings. In addition, the jury can request any additional information needed to be thorough and confident in reaching their verdict.
(U) The types of evidence provided to a jury during the course of a trial include statisticalbased forensic findings, symbolic logic arguments crafted by attorneys for both sides of the case, testimony of experts, and testimony by defendants and witnesses. In a sens~. the judicial system already has had to. deal with a span of evidence much like woqlJt;g~ the case in WMD attribution and has developed a set of standards to convey confidenc~:iiJ":l!Sgregate findings. Furthermore these standards, when met, correlate with a speci:ij9.:Q!Ilge 6t:~J!.t,i:nces that a judge may render.
;::-:::::~..
-
... ...
.,'l!,!f,
(U) Discussing how the judicial uncertainty schema might"fi~ ofus~).~~l
::::'{
characterization of uncertainty and confidence in findings reqt#~. crossf~~:;~e wall between WMD findings and possible US actions - the dreaded policy.. prC"~g~on gtiideline. In US
national and domestic security, the range of responses fQr~giten thf~i:S:and actions against US domestic and foreign interests has been w~J!:9~~~~~~ througii" experience. In the diplomatic area, actions include suspending,diplb"iy~tic relat{C?I]~~,recalling the ambassador, imposing sanctions, and issuing demarc,lt~~ . Intell~~ence acti~ii's include covert operations, covert propaganda, infiltration of colleeii6'ii;!~~y_i~, and various forms of covert influence.
Military responses can include declaring war;:~iif.gical strikes, sending in special operations forces, and complaining through the attache sysf~i!J; . ,~conomic actions could span embargos, cutting off selected trade, economic sancti9ps, andf6bmplaints to the World Bank.
(U) Stepping away from policy;.proscri~~~~ conc~ms for a moment, if.---------., c=:::Jajudgment that it was:~rtJiility ~tertain that the AQ group in an area of Pakistan were responsible for the WMP e\;en~ then_.tjJe President's advisers likely would have .. find and bomb the camp" in the;F;fecb~~ndiifions. These advisers also would have a range of options, per~~ps att:tii.~.way dowl{t'O .. do nothing yet." At the other end of the spectrum, one would e~pCC:t~,1Y;~ry li'iSJ!~Jeve_l of confidence before considering a nuclear attack by the US.
For e~~p category ofWNfll>;:event, from yield-device to minor contamination, the President's adv.i~ers::will c.9tne forward with a range of options, depending on the confidence of the judgme~~~lfl~Y recei\'e If the I
~sessment were weak or consisted of multiple hyp9~~s~, then queries certainly would follow on what additional information is necessary to iiciii~"ihe assessment.
.
~ .. :
(U) As a first step toward establishing a standardized method of expressing uncertainty, we recommend that be exposed tp the existing lexicons on uncertainty. This step can be followed by the group going through a process of identifying aset of standards of proof. These standards ofproofwould be defined to convey uncertainty and confidence consistent with the risks inherent in a plausible range of actions that the President's advisers might recommend. Representatives from the President's advisory entities should be part of this exercise. As with the judicial system, the resulting standards of proof would not define the policy response (or analogously, the sentence in a criminal case). The standards of proof
would represent the fact the there is sufficient confidence in fmdings to support a full range of response options. The President and advisor entities will drive the actual response.
(U) In response to an actual event, I
I
would evaluate the specific hypotheses, levels of confidence, and alternative interpretations presented by the aggregated LE, IC, and TNF
assessments. If the findings are diverse in interpretation, the ~ould ask the community to iterate based on additional information accumulated over a finite period of time.
(U) If a specific hypothesis is not strongly supported by the evidence and one or more alternatives are admissible, /
/should have a practiced operational means for reaching some form of closure, if possible. This could take the form of a pro-side and con-side debate of the supporting evidence and the level of confidence associated with each hypothesis. The IC, LE, and TNF advisers would respond to information requests. If no refinement in fmdings is possible, then the policy community would be informed of the possibiij~es, associated confidence levels, and what information would be needed to resol~i;9f'
significantly enhance the group's findings.
.;:~;~;~!~;::-.
.:
(U) Whether analysis or information assessments originate from the LE;2it;:~~.'JNF .,,,,,.
communities, each implicitly or explicitly works through q1,1~stions ~iii!cal to'~~~ssing confidence. In essence these questions are:
:::''
' '::~:::;..
:~'
-
What is known?
-
What is unknown?
-
What do we think?
-
Are there several hypotheses?
.::::::::~~~~;:: .. .\i;=:=::..'"
-
Is there one hypothesis that is n;19~:c;omp~jllng than 'ihe others?
-
Are there dissenting or contrarY hypotheses?
-
What is the line of argUm.ent for or ag~z _
(U) Using the above as a standard forma(,.~:re~~l~~~~ings 4
I woul~help to convey levels of certainty. Hav\qgthe:3ii0ience's attention and the time to walk through these or equivalent questions :aiia;~s~iated discussions would be the most desirable situation. Uncertainty woul4 b.e:~ommunicated in plain English 'vice a "catch phrase" or statistical statement, in~r~Mion, P.r.ifudividual experience .
.. ::;
~-;:;:=~==~~::: -
.
/~~~!~=:-.
. -.
(U) What ~e~~
((U)j
## (U) Retommendations
(U) Lessons from Previous Studies on Uncertainty
(U) D
published reports on conveying uncertainty;:~*~:has. .p. ~i~li'lar applicability to how might approach ~e problem of_de~elop~g:a meilf~8.f.expre~sinf.
uncertamty.
at report, "Uncertamty Commumcabon;,;ij~ues and GQOd Practtce" ts recommended as a starting point for develoP.ip.g;~~:fJ.t~tlioH::~l~xpressiltg uncertainty inc=]
assessments. The report provides guidan~~:on coitpjriunicaiifi~:;yncertainties and draws upon insights from the literature, from an in!et.#.,ipnal ~;Cperts' wqr:Kshop on uncertainty communications, and from several uncemiiii~~~xperiments. Especially significant is that this report focuses on communication to non-teclli:HCitl audiences, with special attention to policy.:
makers.
-:::~j~~;/
(U) The report expresses severatpP.po~f;rinciples. First is understanding that the target audience may be under extr~J,riif'iWie qonstraints and generally will have non-technical backgrounds. These facts fq~tilate t~e overarching character of the communication, which consists of a primary l~x~r, ~Ji'~~~ -~~J6ottom line assessments are presented, and a secondary layer, where the detru!e.<f data'~~~logic are presented. Most communication time will be spent on tlt~.~~l~y~r. If a written format is.used, the audience will typically self-select the primary'i'il.:Li.t;to reaci:J:si9.PPing much of the secondary layer. Consequently, colllJP.QQicatio~ipf unce~ty must be done in this primary layer; otherwise it may not ever
. besee:ri1b'x.the:target audience.
-~.:~~:~;~
~:-:;-:-._
(U) A secorid';P.:r;iAciple is that the audience will process the uncertainty within their personal frame of refer~ii~e and according to their biases and heuristics. The report clearly emphasizes that non-technical audiences will tend to relate better to verbal expressions of uncertainty than numerical expressions. Thus, there is some risk that qualitative expressions could lead to different interpretations by different people. This principal is discussed extensively in the report.
(U) For the audiences to make sense of the uncertainties, it helps if they understand how the assessment was conducted. It is not merely a matter of reporting the uncertainties themselves, but the uncertainties also need to be properly reflected in the formulation of the main messages that are conveyed. Specifically, when communicating uncertainties to the policy-maker, the following items should be addressed:
-
Reporting types of uncertainties and how they propagate to the outcomes.
-
How uncertainty was dealt with in the analysis.
-
Implications of uncertainties.
(U) The report identifies some important considerations in communicating the main message.
Specifically:
-
As assessments evolve, explain any inconsistencies with prior assessments.
-
Report minority views and clearly label them as such.
-
State the essential conclusions in a clear and concise form.
-
Do not make statements that you cannot back up to a satisfactory level.
-
Aim for policy-relevant conclusions
-
Integrate uncertainty information into the formulation of the main message.
-
Explain that additional information may not always reduce the uncertaintY-~~ ,/
-
Explain that assessments could change as new information becomes avail~bt~:;~;
## (U) Organically Grow A Lexicon
(U) To enable rapid and credible attribution among the primary communities of intelligence, law enforcement, and technical nuclear forensics, a common lexicon is required to ensure that key conclusions with their associated levels of confidence are properly formulated among the communities and conveyed to the decision-makers. Presently, a common lexicon does not exist among the primary communities. Each community has developed its own way of conveying conclusions with their corresponding confidence levels. Based on discussions at SHARP 2009 and taking cues from lexicons arising from massive social networks, we suggest a method to implement a nuclear attribution lexicon.
'
.
(U) Recent history has demonstrated successes of diffusing language change iR~J~ge-scale social networks. For example, approximately 10 years ago, terms such as ''w~liii~.:
-
0
"browser," "wiki," "blo~," "tweets," "broadband," or ''weblink" were D;?-!~~-widel~~~~~
across all US demographtcs as they are today. Today, these new termstllave:C:fiffused across much of the US demographics mainly due ~o a large segm~I].~ of the J?9P.!.Jlati~fi:f~;,social network) that is engaged in internet-related activities. Theo'ffiffu,sioii sp~~-of new language, which is lexicon-related, is proportional to the size of the netw,<!~~:, Res~~has shown that regular and small-world networks will show gradual di~iQn iti':lab.~age:'thanges, while random and scale-free networks exhibit rapid diffusion}, Sihtilarly}i).~m;ork influences aid in the implementation of a lexicon among s~~~~<;Q~tftijij~s.
/
0
~~ =~ :-:~::;:;;
.
:~~~::~..
.
(U) Network factors are dominant in orgaiD~Iy 3J~ing ~~~~on lexicon among separate communities. For example, for a massivl-~~9-!~!.rittwork, su~h as World ofWarcraft, the common lexicon was grown by large numbeii;~q~.persons frequently using specific terms with specific meanings. In time, the lexicon is devdb'~:~d implemented by the sheer number of people (i.e. millions) using and acceptii].g:theseJeims and their meanings within their network. However, the nuclear attribuqQtfcomoi\mities do not have a massive social network to drive the use and ~ceq,~c~~of a common lexicon. Instead, a similar effect of organically growing a conu:p'on _lesticol! can be accomplished by increasing the frequency of interactions among the ~~mli\uiiities ~ing specific terms. Research has shown that children's vocabulary iii&~~;~~i:ically with the number of interactions (i.e. network size). 5 Sim!].ij'ly, ~~~t~.?.'icon is "grown by implementing a large number of person interactions using sp,~ific::i~htis, otl:~~-~a~,is not possible, implementing a large frequency of person interactions in afsinaller iietwork to achieve the desired effect.
_ ... -:;~;:;~::~
_:::-:
)~
(U) To d~~~J;6.p and implement a common nuclear attribution lexicon among the primary communitie's'~l~~ Jrequency of person interactions using the specific terms would need to be increased. At'P.resent, the primary nuclear attribution communities of technical nuclear forensics, law'~nforcement, and intelligence embark on average bi-annual or longer time interval exercises, which are too infrequent to simulate the number of interactions used by a massive social network in generating a successful lexicon.
(U) From a sociolinguistic perspective, the development of specialized language or "lexicon"
serves two primary purposes: to allow the group members to communicate clearly with each other, often about specialized concepts and topics to a degree beyond what everyday language facilitates, and to identify in-group vs. out-group members. 6 That is; evolution of language functions as a key indicator of.formation of a group. The degree to which a common lexicon emerges from these interactions is a measure of success for genuine community.-building within the nuclear attribution arena.
## (U) Re(::Ommendations
-
(U) The following recommendations are presented to enable implementation of a
common nuclear attribution lexicon among the primary communities and are based on the
demonstrated success of lexicons developed in massive social networks:
-
(U) Since the primary nuclear attribution communities are small and crossed, much more
frequent interactions among the communities are recommended to create the similar
effect of developing a common lexicon with many interactions in a massive social
network.
-
(U) The increased frequency inte~tions among the communities could take the form of
monthly or bi-monthly secure, web-based, virtually-linked mini-exercises:;~n.a,.particular
facet of the attribution process to calibrate the communities with the speci~~~{~rms and
their meaning.
.
;; ':;:;:;.,
.... 0!'-!l~tlo
.~
-
(U) The mini-exercises should be brief(e.g. -2 hours) to foster and sq!g~.develb"~p;1~ht
and acceptance of a common lexicon and to minimize disruptions t~{ong~1,pg_ mission
areas.
.~:::r
===:!;:.
~~::~t;=~~:
-
(U) Promulgate a mechanism for ongoing communication llJ).d ihte~tipn, iri' addition to
the mini-exercises, to further facilitate development of a cpfiiWpn lexi~r{ indicative of a
growing community. Virtual means of interaction may::~ b~si;~%~~~;'as community
members are geographically dispersed.
;;:.,;:
-~~~-
## (U) Split Second Decision-Making: A Law Enforcement Perspective
(U//FdNO) In a nuclear/radiological attribution scenario impacting the United States, it is the US Pr~~nt's responsibility to decide what the US strategic response will entail. As the President may have limited insights on the capabilities and limitations of Law Enforcement
(LE), the Intelligence Community (IC), and technical nuclear forensics (TNF) to resolve attribution questions!
Jmust provide each new President (and his senior advisers) an understanding of the capabilities and limitations to ensure informed decision-making.
## (U) Csi ~((~~. J~,~~Eived Misconceptions (U) ~~C:J~~Isioiii:Ilii~~. And The Public May Have Preconceived Misconceptions
reg~d.ig.g the tYf;e and degree of certainty of evidence available to support WMD attribution assesslri~~-::The crimin3.1 justice system has been challenged by this issue, as jurors have been exp6~<to contemporary television crime shows (e.g., Crime Scene Investigation, also known as csf)~j1~d ~ften ex~ec~ irrefuta~le _eviden~ to have t>et:n ~blaine<~: an~d presented for ::p.;.cnmmal convtction, all wtthm 50 mmutes. In a stmdar fashton, ~
and supporting anillytical personnel may be confronted by unrealistic expectations from an alarmed public and an eager body of elected officials, all of whose concerns will be fueled by an aggressive media that may further fuel these misconceptions. In LE circles, this phenomenon is referred to as the "CSI Effect." It is important to convey to decision-makers that attribution and investigative speed often are constrained by the laws of science, in addition to adhering to prudent plaiming. 7
(U) Mitigating the "CSI Effect"
(U) "Max Houck of West Virginia University complained of the 'CSI Effect' that has pushed expectations of crime scene inves~gations far beyond what is achievable."8 "Jurors now expect us to have a DNA test for just a~e_:~~~case. They expect us to have the most SEC~FORN
.
advanced technology possible, and they expect it to look like it does on television. "8 This same level of expectation will influence decision-makers evaluating nuclear attribution assessments. It is important for the contributors to to be cognizant of the unrealistic expectations of their customers, an!i develop ways to operate effectively in this environment.
The LE community has experience that can be translated into the nuclear attribution environment. For example, here are some standard practices:
-
(U) Know your capabilities. Conduct a self assessment of your organization's
capabilities. Evaluate your capabilities. Determine how much time you need to conduct
the appropriate testing or assessment.
- (U) Educate your partner( s) in an investigation on the capabilities and limi!&tigns of your
organization.
:;;.;.:
(U)D customers may be reluctant to accept that attribution answers IJliJ.Y.~:=~~~is:~:::::
unresolved for years, or possibly indefinitely. It is plausible that the re~i1ii"8t;~ asses~ment of all available evidence and sources may not result in a defi.nitive ~wer on th~~pei:petrator
" :~>
.. k .,'!, .. ~..
\~, and complicit actors.
.-.
:::!--..
.''
(U) LERC: Risk-of Being Wrong
-
(U) a~s~ upon collaborative or high-confidence intelligence, provide a clear analytical
assessnie"#b~J?.ere the findings support a single outcome.
Or,
~:::.'
-
(U) Provide an alternative analytical assessment to the majority opinion, based upon
inferential analysis or intelligence that may presently weakly support an alternative
scenario, so that this view will not be lost as new information is gathered.
(U) The analyst/operator needs to be empowered to make decisions with a degree of built-in flexibility to enable, for example, the development of alternative assessments, and needs to be held accountable for his/her results. Simultaneously, the analyst/operator needs to be encouraged to think outside the box by operating without concern for punitive consequences.
(U) Empowering Personnel
## (U) Minefields And Mind Traps In The Attribution Problem
(
It is charged with providing the best possible attribution assessment, under tremendous pressure and in the face of uncertainty, to provide key decision-makers with information that is accurate, reliable, and actionable. In order to effectively and accurately perform under these conditions, there are a number of considerations from a social-psychological perspective that must be addressed.
o~~ly spf:aking, in order to perform this most critical task, it will be es~~nti~ILI _
__j
: ~
!~~~~~d h=~~~i~r~:::;~ ;:~:~~~U::;e::.:-I::r::;;rks .. :.:~.:::-::;;~~~~!;~;;,;.!~
-
(U) Establish sound analytic processes, procedures, and methodologi~8-;:::!~:..
-
(U) Develop the ability to craft and deliver message eff~qtively ;;;:;{
-:::;:;;/'
-
(U) Adapt to and function effectively in a crisis enviroiliilen~ :
-:~::;;;.
. -:.
.
.
.:;~==~-
~ .. !~f.:;; .. ;:
(U) The following articles discuss the importance of each 9f.:tb~~~~~!~J:.llel1~~ the potential obstacles to achieving success with each of them, and rr..plnmeridati~~s'for how to overcome, or at least minimize, these obstacl~~::~!~M~~~ilHl~~~~ by addressing each of the following main questions:
~/ ~:;:;=:
:;;::::~-...
1. (U) How can attribution analysis b~:i~~~X.~ ;:;::
:3::
by taking into account
various psychological or social factors? ';;~~:=:.
2. (U) What sensitivities or constraints on reai~i:if~:pt>licy-makers should I
~e aware
of or sensitive to in order to improve .~-effecfiveness of our communications and the
value of our attributions?
. *::=*
3. (U) How can 1
lfu.nttt:~~;be~;r across the three attribution communities in terms of
information sharing, Iatehttc6mm\inication, and coordination?
## (U) The Road To Hell Is Paved With Normative Cognitions
(U) Accounting for Psychological and Social Factors that Degrade Response or Inhibit Information Sharing in c=JAnalysis
(U) The judgments and decisions.ofth,__.------'-'-==--.:=-====:....=..-=:.;:;
that influence everyday decision-making.
~~~=r.~======~~
:. ..::
-: .. :-:
(U) Core Concepts in Judgment and Decision-Making
~:::;;;:::..
.
(U) Given a choice between two or more options, whether related to min9.r:d.~cisig~zQ\":fhose of great import, like attribution analysis, how do we decide between th~;iiililifpJc;: opti()its and their relative merits? At some level, the decision-making gr;gcess aelte.~ strai'~tforward:
I '
;1 '!iJ
'
we consider those facts that we feel are relevant, weigh them in .termso.~~tpportance, and reach a conclusion. In many ways, however, this is a "black bP.~~~procesS::i!!-:which the relevant facts get poured in and an answer emerges. In thi.!i~ktlfi~;we unpack the "black box" and describe some of its internal processes that re~:t!l~'m the erii~fg~nce of judgments and decisions. We will also explore the implic\ltiP.IJS oJ:ili-ese normative processes on the decisions and judgments of adversaries, I . -' - ~d p~lici.~:J~~~ers .
.
~-:;:~
;.:1!~
~:::,
, .. ,a .. ::~:;:-..
3"'
/"
(U) Rationality: Bounded and Otherwise -:;::~;;{
(U) Part of the attributiqn analysis will involv~-~~~iiig the likelihood that a given group or
individual would engage in actions of con,;9.!=rn ( e.g;=smuggling or selling nuclear materials, or
carrying out an .attack of a given type). ;-Such an analysis often begins with observations as to
whether it would be rational (ot;:fli~indiV:idual or group to engage in that behavior. Analysts
must bear in mind that assuP.i"pti_oos of:rationality or irrationality can be hazardous.
Assessments ofrationalicylifi_:atlonalitY~are generally made on the basis of the extent to which
a given behavior is c~qsisterit~~~:s~tietal/cultural norms (related to religious beliefs, moral
values, rep~&\~ce;:~~-), arises from a decision-making process that follows a logical
syllogi~w(of.:i~;~~nsiSi~t;~igt what observers would view as providing favorable outcomes.
(U)"ti~~~~z~~J=:f ass~; .rationality or irrationality are great, for the following reasons.
First, analy~j~ of the behavior of others from one social or cultural perspective may be invalid
if it does n6't!iake into account the social and cultural standards of the actor. Second,
rationality in ;~frather than theoretical, terms is context-determined. The fact that a given
behavior may be considered aberrant or irrational by the majority of people does not render it
irrational on the part of the actor behaving within the context of the actor's experience. For
example, individuals and groups can be behaving rationally even if the behavior is based
upon misinformation or beliefs that are inconsistent with reality.
(U) .Finally, utility-based approaches to rationality have been displaced by the concept of
bounded rationality. 10 Taking into account that human beings have flawed memories and
limited computational skills, Simon's theory \Vas based on the notion that -we use mental
shortcuts and rules of thumb (see biases and heuristics, below) to allow us to compensate for
our cognitive limitations. In developing the theory of bounded rationality, Simon described
decision-making as a process of searching for alternative behaviors and choices that will
## Sec~Oforn
result in the actor reaching his or her goals. It is a search process in which decision alternatives are discovered until a satisfactory alternative is found. Simon coined the term
"satisficing" to describe this process: a choice that may not be optimal, but sufficiently fulfills the criteria necessary for reaching the actor's goals. This comes into play when judging the likelihood that a given party will become engaged in a certain activity. For example, an adversary faced with a range of options may choose one that may not appear optimal to those analyzing the behavior, but that sufficiently meets the adversary's criteria for some degree of success.
(U) Biases and Heuristics
.
(U) AnalystsJ
I those individuals and groups whose behavior they ~.ana,Iyzing, as well as those who will determine a course of action on the basis of the attribution" assessment, make decisions based on heuristics, or rules of thumb, and biases. 11 As noted;~~~xe, th~.se are the mental shortcuts related to bounded rationality.
..::::::::
:~:~*;;;:
::::
SUIIIlll01ized by s.-m include the following:;:: ,~:;:"/'
5
'~:~~~iV . ,
/~~~;~ :'
".;_~~;~:~~
.. - )i.'
( o) Extremeness Aversion: Given a range of altemative~;fi;om which~t~f.'(~hoose, people avoid extremes, which give rise to compromise eff~~~-~:~af~~~~nsidered' to be extreme is influenced by the frame in which the alte~uvesf#:preseii~9. )?or example, in merchandise purchasing decisions, an individual pre~~~ witli:i\vo simii31-Xtems is more inclined to purchase the less expensive item ofthe'tw'i:::;ijo_wever, this individual will tend to choose the more expensive of these when presented witii~~;~ird similar item at an even higher price.
The same behavior would be expected to bold ~e;:w.l;len individuals are presented with items related to courses of action or levels of ce~ty. "tit the context of attribution assessment, an outlier opinion offered by one member.~gitlie gro~p may lead others to agree to a position that is more extreme than th~!.1ni~!?ha~e initially chosen.
(U) Hindsight Bias: If ap_.e~~p.{~c~~;there may be a tendency to believe that the event was inevitable. This in tl!m.::fiifltic;"ri.~~~tJi~extent to which the same event is viewed as likely to
be repeated. This c!irr~pt relate~'"to the notion of "Black Swans," where low probability, high impact everi~:;iihreiro~~ct are believed to have been predictable. 12 Taleb argues that acts of terrorisfii, lik~'~iiifts in tll~~~liancial markets, are "Black Swans" that are in fact not
.
predl~~~-!~ ~~fhtust be aiiticipated.
'"~~::!:~:,.
(U) Opti~~~~~:~~~: Sunstein suggests that human beings tend to be optimistic. Individuals and groups inv:Qlved in attribution assessment, as a result of selection bias, are less likely to be prone to optimism. However, optimism bias may play a role in how an individual views the validity of his or her own decision-making methodologies and accuracy of his or her assessments and judgments. The most difficult assessment I
~ustomers to accept is that no answer may exist, which may conflict with an individual or group's optimistic bias. It also plays a role in bow the public perceives risk of harm from illness and disasters.
(U) Overconfidence Bias: This refers to the phenomenon in which people, including almost all professionals, tend typically to have too much confidence in their own judgments, most likely due to insensitivity to the weakness of their underlying assumptions. 13
(U) Status Quo Bias: Human beings tend to favor the status quo and require considerable incentive to depart from it. This relates to choices involving changes in behavior, including changes in decision-making strategies.
(U) Confirmation Bias: Believing is seeing. In addition to those discussed by Sunstein, confirmation bias is a major consideration I
I The psychological research is replete with demonstrations that context attu trclmmg onectly influence how we perceive objects and events, from optical illusions to the behavior of individuals and nations. Modem geopolitics provides us with examples of leaders who interpreted events according to their preconceived beliefs about the groups or countries involved in those events, rather than upon actual data, e.g. the adamant belief by some adminis~tion officials that Iraq possessed WMD in spite oflimited certainty on the part of the IC .
.(U) Confirmation bias, also known as ''belief perseverance" and "selective pe~~~~tion," is a concept that captures the fact that ~pie tend to select and interpret informatiobl~~-~-way that supports their existing worldview. 4 This concept applies to how thorougW,:x. a persQ.)t~ll read a report, i.e., a person who disagrees with the conclusion of a repo.rj;iiiliY:~nc,>t rea'd''it at all, may read it superficially, or conversely, may read it in ~~t dep~,Vt an eft6f,t~ find fault
"thth l .
-'
~.
wt e cone us1ons.
-.
.;~. .
:;~\:~,,. ' ,:-
.<~; ~ -~.
-~~: ..
(U) 'The CSI Effect: An additional source of bias may aris~Jronf~?Spectations regarding the type and level of certainty of evidence provided by ana\Y..s~'io 1
r decision-makers, and the public. The criminal justice system h~,!?~!l,fn?ni~;Jhis issue lis jurors, having been exposed to modem television crime shows1,.oft~ff:~*pect ~fu!a~le evidence that will have been obtained, analyzed, and presented fQi;;:a critniAiii convictf8Ii; all within 50 minutes.
"Jurors now expect us to have a DNAtfsf:i"ri~:j:us.t~bout ever case. They expect us to have the most advanced technology possible, and:tlt~t"expect it to look like it does on television." 7
... ,. .. ~.'!~>
(U) Similarly,!
lthe decision-~:~er~-.~,.,_..
may have unrealistic expectations about the level.~~certainty of the information that analysts can provide, as well as the time fi)wi~~W. wiiich it can be provided. Each of the three communities contributing tq::th~-~sessptent process needs to be prepared to deal with this environment of unrealistic exp<ftatio~. How can this be done?
::::::..
":a:~~~:: .. :~:~=~~::;::~
-
(U).~>W ~!k~stom~::jr-:~"-----------,lcontributing communities should establish
~.~app~~;:with'~~.~~lti~te customer. This will include establishing an agreed-upon
_.::il'exicoti"~4;' ensurt(~fmessages are clearly understood. This will be especially critical
.. -:~::;:*hen cainmunicatiil.g technical information to a decision-maker who has never been
. ~;i{&~ed to such information.
-
(Ufliip;w_the capabilities of each analytic community. Each of the contributing
comni~ities should conduct a self assessment of its capabilities and limitations.
-
(U) Educate the customer as to the capabilities and limitations of each contributing
community.
(U) The most difficult assessment I
Customers to accept is that no answer may exist.
The outcome from the assessment of all the evidence and sources may be that a defmitive answer is not achievable.
## (U) Heuristics
(U) In addition to biases, decision-making is influenced by heuristics. Heuristics are "rules of thumb" that help speed the decision-making process based on past experience and
## Sec~/Noforn
knowledge, both individual and institutional. They have been referred to as part of the
"Adaptive Toolbox" of decision-making. 15 Sunstein describes the following heuristics:
(U) Availability: When an incident involving risk is active in an individual's memory, as a result of either severity or length of time since t:b:e occurrence, there is a tendency to overestimate the probability ofthat risk being realized. Probability of risk estimates gradually decrease over time. For example, popular estimates of a terronst attack were extremely high immediately after 9/ll, but have decreased over time, regardless of the actual geopolitical situation. Similarly, estimates of the likelihood of a workplace violence incident increase when such an incident has occurred within a similar community.
(U) Anchoring: Initial choices and probability judgments tend to serve as anc~ors _in the decision-making process, even if they were made on the basis of imperfect infpJ;rtiation .
. .
~-.
Anchoring will take place early in the thought process and is a very powerfuHout~~ ofl!ias in thinking and can thus prevent an accurate attribution. Anchoring pre':~~ us fro~ls.e,ing or accepting new information when it does not fit the world view we hc;>fd'(i~~-~~our '-""
preconceived notions of who did it, who supplied it). As a.r.~sult, it.~~:J!nportiffi~:l9at analysts and decision-makers be aware of the weighting of initial jtiElgments'ino(der to be able to more fully incorporate new intelligence into the decision-ma.IQrj~,JlroceS~;~::::-
.,, .. _~ .. ~-..
. ......
(U) Case-based Decisions: When faced with alternative,.ctiJi~es :1~~~~ value, people tend to reason on the basis of prior cases (i.e., .~Q~.js tlll~iiJ.~w case siihilar to or different from a similar prior case that was analyzed)/Tiil&:f~p.ds tO:;ljjnit ~reativity and the ability to incorpomte novel information in the anaJtl.c proc~~-
MoreaYff~, use of historical analogies
(e.g., this will be another Vietnam or MWi'~~~-~e and anchor people into more rigid ways of analyzing data, resulting in significant distQ~~-~s and biasing.
~-::~:~::;;
~~~
_ .. :,;."
i/ .. ~
(U) Bounded Willpower
.;;::
(U) Bounded willpower is a c~ri~~t th~t recognizes human beings' desire for immediate gratification and reward. It ;SPe~~ to t!J.e issue of impulsiveness, which in the case of attribution analysis and. ~se$~ment can be thought of as a rush to judgment or premature closure. Once recogpi~ed, b()1W~~~illpower can be managed through a variety of mechanisms.,.iaciuQiiig,criticat"re~lew by colleagues and members of other groups, as well as other che~~;~.biilali&~s. in $e decision-making process.
.. :~~::~
-~~:~;;
~~~~;~:~-~~
## Ruj;;Ti~D~~Lf-Lnter;;~
(U) In addi~i6~:t~.bounded mtionality and bounded willpower, bounded self-interest is another factor"t;Jiat contributes to human behavior and disproves the notion that human beings pursue behaviors based purely on maximization of personal utility. People care about other people, and about causes and values, leading them to act in ways that may actually harm themselves as individuals, yet provide some other reward. This comes into play in at least two forms in the attribution analysis process. First, it applies to the analysis of the likely behavior of adversaries. Second, regarding the idea of attribution analysis as negotiation, whether an individual feels he or she is being treated fairly or unfairly in a bargaining process has an influence on the toughness of the stance that is taken by that individual in the ongoing bargaining process. As attribution analysis is a collaborative process, it sneaks to the importance of establishing a culture of openness, respect, and objectivityLI -----'
(U) Implicit vs. Explicit Decision-Making
(U) Analysts will make explicit decisions about what information is and is not important on the basis of their rofessional ex rtise.
The decision as to the relative importance and mearung o a particu ar p1ece o 1 otmat10n reflects the professional judgment I
I influenced by factors such as time pressure, severity of the threat, and recent social and political events, both foreign and domestic. These decisions, like explicit memories, can be explained by reference to specific data points and events (i.e., "I know that this is Cs-137 because the following analyses were run yielding the following results.")
(U) Decision-making, like memory, also occurs at the implicit level. lmplicit:Ut~tpory, which is also called process memory, cannot be explained by reference to known sp~~i~~ prior experiences. Similarly, a host of other factors ofwhich the decision-maker is:n~t;~:Qpsci9usly aware can influence the decision-making process. These factors can be !JD..ig~e to ffi~;~;:;i~
individual, who is reacting almost instantaneously and in light of pas~ ~#~en~~~~s ~d
information, heuristics/biases, and subtle clues of agreemeP.t;,or di~~~!.llent fiQ:@~others engaged in the group decision-making process. Psychologi'cid ~xperiiii~#-t$. haveshown that over time, gist information, i.e., a general sense of the meaning~~~!!: data"~;:i more persistent in memory and has a greater impact on behavio~;,t9an vew~tin:a, ihformation. For example, information that a particular radical group has.;9.tm involve~~fu. attempts to traffic nuclear materials (although has never done S'Hlis;c.~ssfiill&;) .. .will persist in memory far longer
.
- .... re .. ..
.._. po than the details of the failed trafficking att~mpts:-;:::~::
::~;:;;..
_,_.;:=:~..
~:~:~
l:::!M
(U) Implicit decision-making, like clirii~~f~~gqt~rit in medi~ine, is a form of intuition.
Based on a combination of explicit knowledg~~4 gist memory, it provides an informational infrastructure for decision-making of which the'd6~i!!ion-maker may not even be aware.
Tpese are essential tools and add to the q~ty of~&ision-making in repetitive situations
(i.e., conditions or events with which t:IJ~:'decision-maker is familiar). However, when dealing with unique situation;;-ot:\Y.he~ifattempts are being made to deceive the decisionmaker, they may actually hi;tideJ;".accllf4te assessments.
. .
~~~ ..
;::
(U) With regard to d!e:~~flu~~~i~~:~~ opinions of other members of a group deliberative process, in~ixip~al~~i:py in terms of their ability to perceive and the extent to which they are influens"by:\ij~. reacti~~pf.,others. This is influenced by individual personality factors and life e~~rienc~~~;Which irt'~tj;itfi may be tied to such specific individual factors as variations in the-:tiliib~~n.it}g"ofmirron\eurons in the brain. In addition to playing a role in anticipating the physical~~yiors of others, mirror neurons are believed to play a role in the capacity for empathy, oili~~.e known as the understanding of another person's emotional response to a given situation~:;in the high stress setting ofl kieliberative process, a presenting analyst can be'expected to be constantly reacting to the emotional cues of his or her audience.
The ideal analyst would be one who has the capacity to incorporate both the data and more subtle cues provided from others, without his or her objective analytic capacities becoming overwhelmed.
(U) The Hazards of Bias, Heuristics, and Intuition
(U) Biases, heuristics, and intuition all have an impact on normative decision-making. As noted above, biases influence the manner and extent to which information is processed and treated. Heuristics have been referred to as part of the Adaptive Toolbox of decision-making and can speed the process and increase accuracy. They are especially helpful in situations where the problems and tasks are repetitive or do not vary widely, and where active attempts are not being made to deceive the analyst or decision-maker. Where such attempts are being made, as may be the case in events coming to the attention j lor where the problems being analyzed are unique, these factors can have a negative impact on the accuracy of decision-making.
(U) The tasks of analysts serving j lwm be repetitive to some extent. This certainly will be true when it comes to analyzing forensic data. The attribution analysis itself, however, is less likely to be repetitive, especially in the case of major events. These events will be rare, but potentially devastating in their consequences. They fulfill the first two characteristics of Black Swans, and in hindsight many would attribute the third quality as well, that the occurrence was predictable. With low incidence phenomena, th,c;:re j~
considerable risk in resorting to shortcuts that derive their validity from the si@J;hity of current and past situations and the accuracy of the underlying data. In such sihl~~ons, the Adaptive Toolbox may become the "Maladaptive Toolbox," leading to f~.~,,foncl~iR~:
reached with high confidence due to the explicit and implicit knowledg~:ffia'~~~.!?-~es wiih expertise, overconfidence bias, and confirmation bias. Thi.s, is partie:(J.J*ly prot?ler;natic where the perpetrator has taken steps to deceive investigat<i~ in ord~'t::ffi:~~scape.:ii~tection and identification. As such, factors that provide shortcuts for decisiQn.,-maidfi~::must be actively guarded against, and a careful balance struck between the ':!~i~izati~~:!?f bi~~s. heuristics, and intuition, and recognition that!
fvill be facing aJ~!ick Swan:~4ttially every time it is called to action.
. . . .
.;::;:::~.
:
(U) Careful crafting of the message and.~q~~~:::i~n-~f:!~;~~ihod of communication
(including the communication of uncertaiii~)..!=a..fi:bontribute greatly towards combating the negative effects of biases, heuristics, and inhiit;i'9n. A detailed discussion of these techniques is found in the section of this report entitled, M~~~RESSING UNCERTAINTIES IN
CRISES.
.:~>
-~~;:
_ .. :::.: ..
...... ~~ ...... ... . :::: ....
(U) Beliefs *Are* Possessions,/;.)_:::~
~
(U) One challenge to tl1t:.k~<t'~fthin}()ng-outside-of-the-box that attribution questions may require is how we de!ll:W.th eiiJ!t:t:::ft1' the pre-existing beliefs on the part of policy-makers
( confirmatiop, \>ias ~::~la~d to eiili~r the correctness of prior policy approaches or interpretatio~~~:}fthe'rii~J!yations of potential perpetrators), or (2) the prior institutional apprqhes thS:tiprevio~l};,iliave been adopted to deal with the problem or issue (optimistic bias-~::r~J~ted;fri previous'sOPs or the historical manner in which the relevant organizations have ini~~~~d or handled these issues). Once beliefs have been adopted by policy-makers or implemen~~~Y.-organiZ;ations, they become possessions to cherish, protect against change, and adhere to -~~en in the face of considerable evidence of failure or inadequacy vis-a-vis the problem. For.a'ttribution, it is important for policy-makers and organizations to avoid the trap of having these prior beliefs function unduly as anchors that prevent alternative or competing analysis from being considered. This is especially important, given the high policy consequences likely to flow from any assignment of WMD attribution.
(U) Anchoring and the Pressure to Round up the Usual Suspects
(U) There will almost certainly be a disconnect between the speed at which the national leadership must respond to the policy/political environment and the slower pace at which forensic evidence, technical analysis, and law enforcement investigations can proceed. This gives rise to an anchoring problem (i.e. a tendency to anchor on the usual suspects in
## Se~Orn
attributing responsibility for an event). Given the magnitude of the likely national response roany substantial WMD event, those involved in the attribution process need to be cautious of leaping to conclus~ons ahead of the evidence, based on the emotions of the moment and the general tendency of people under stress to rely more on biases and available simplifying heuristics (such as stereotypes, ideology, pre-conceived belief structures, etc.) Two examples of this are provided by the Oklahoma City Bombing and tile Centennial Park Bombing at the Atlanta Olympics. In the former, there was a rush to judgment that the perpetrator was a Middle Eastern male. In the latter, a security guard was falsely targeted as the perpetrator because he fit a supposed profile of those who would engage in such acts.
(U) Moreover, given the tendency toward anchoring and reliance upol simnlifvinsz heuristics, especially during times of heightened stress and crisis, it is important_
~o ensure that its ongoing presentation of the evidence does not unduly bias policy-mak~t:.s{ This is particularly important because the evidence may be relatively incomplete duriii~~~ tim~
rra.ne when some retaliatory decisions of great consequence may be um:!~t:.SY,ruiid~~ti,9#
:~~~,~-:01:~-'
~~-'!
(U) In addition, related to this anchoring problem is the r~9~nition.~,lr----1~~ysis should avoid assuming or describing terror groups or nationalleadeis from'hos~te8' as being crazy or irrational, since this is almost invariably inaccurate (b~.~ on o'ifr~9\Vfi affect or emotion) and distorts our own analysis of their current an<U\lfui-e:~~paviof.:' At worst, it creates situations of self-deterrence on our own part or l)~gtl1y inacc~te" predictions of what futUre behavior they may pursue against us. 1~~.1!!~~ 'Y-itii~~f exceptioh, national or terror group leaders are rational if you understan~i:;tlie"cd~~~xt witJiiA.~~ich they operate and their own belief systems. Often, describing t!t~~ indiv~t}uals as cijey reflects a normative judgment, that by Western standards, t:Ileiii~e.l}lly,io"r is irrational. While that is perhaps so, leadership analysts routinely note that undetstQQQ. within their own country, regional, or group contexts, these supposedly crazy or iLtf!iiiiaUeaders are perfectly rational and predictable. Since a potentially critical tasl<.:_
I may well be ascertaining. motives among potential perpetrators, the likeliJ?,gp<i of their involvement, or the credibility of possible follow-on attacks, it J~~portiiit to adopt a more objective analytic frame in assessing these individuals .. Anl:!hi;iing~to stereotypes of opponents is analysis by labeling, and distorts our.unders~di!u~~'Ofthe.~~l threat environment.
~i
.:::: <:::::::-::=:=
(U) How We Perceive Risk
(U) The path-breaking work in the field of risk perception by Kahneman and Tversky has resulted in the development of prospect theory, which argues that choices are strongly influenced by whether the problem is framed in terms of potential gains or losses, with people being more risk averse in the domain of gains and more risk accepting in the domain of losses. In the nuclear deterrence context, for example, prospect theory suggests policymakers placed in the position of obtaining potential gains (i.e. conquering a neighboring
## S~Rn
province or country, making a preemptive strike to disarm the retaliatory forces of their opponent, etc.), at the potential cost of the loss of their own current resources such as national survival or military and economic capabilities, would be expected to be highly risk averse. In contrast, policy-makers facing a situation framed as one in. which they were facing pot~ntial losses (such as being invaded and conquered by another state, or facing destruction of their political regime or economic capabilities), would be expected to be highly risk accepting. In other words, prospect theory clearly shows that people are more willing to take risks to avoid losses than they are to obtain potential gains. A good illustration of this work is found in their book, "Judgment under Uncertainty: Heuristics and Bias."16
(U)I
I
this is relevant to any discussion of how issues of uncertainty regarding intelligence or other kinds of assessments are communicated or considered by,.pol!~y-makers.
According to prospect theory, we must remain aware that attributions (or the {1~~-t~Y responses likely to be called for by given types of attributions) that are framed in terms oilc;i~~~s wjll elicit far greater willingness to accept risk by decision-makers than will_~J~-~utiotis~~ed in terms of gains.
/ t:~.
. .
::~:; . *.:=\t;;b_.* ==~:~~~::
(U) Hot Cognition
,;;=>~
::::: -
(U) In many ofthe contexts in which~will activa!~~:e;sp~~t~l!' cas~~ of postdetonation or large-scale RDD incidelits,tliee'motions gf.~ruysts ail~:JX}licy-makers alike will no doubt be running high. Hot cognition.(~f{e~t or:~ptions) ha8.an immense potential for distorting our perceptions of the enviro,P.riiiririi!i~ ho~~-w~i:i.nt~rpret information. It leads us to more extreme judgments of info~fi~n, perijilps far trey:(;ii.d what they warrant. And it may lead us to fill in the gaps of missing "o(w~jgiious information with emotional filler that could seriously distort our assessments. Higli::4~grees of affect are well-known for creating a polarizing effect on judgments and perception5;:~M!~g them toward whatever emotion
(positive or negative) that exists until thex.~ach ~b'far ends of the continuum. The emotional response from the public anq._p<mcy-makers in a WMD event is likely to'l ~~..., I
b produce information fa:st~tthan" it becomes available, to reach attributions more quickly than is warranted, apd tg.follow along with certain attributions of blame in the heat of that J;lloment. It is impq~f;~atl lbe aware of this hot cognition component, as it likely will be involv~d}tl"alf~~:~enario_s,. I
lwill need to structure itself to insulate itsel(~d ttiih.hnize the effects of hot cognttmn on its attribution efforts.
~r':;~ ~;~::::=::
6
:pon bolli an&ysm and policy-~ md how these migp,ta;ffect attribution analysis also requires consideration. For example, individuals unijif~high levels of stress in ambiguous information environments tend to rely heavily upon stereotypes, analogies, or pre-existing beliefs, rather than upon the information at hand, due to a perceived lack of time to reach decisions. Thus, in addition to the anchoring problem discussed earlier, high stress levels can serve to make these anchoring effects even more powerful. In addition, traditional group malfunctions associated with groupthink are seen as being partially triggered by high levels of stress in these groups. Taken in conjunction with the contraction of authority problem and the tendency of groups under pressure to have a collapsed time perspective, groups have a propensity to rush to solutions, have premature closure of debate, and bolster their pre-existing views, rather than challenge them during crisis. Janis and Mann's book, ''Decision-Making: A Psychological Analysis of Conflict, Choice, and Commitment," provides a solid overview of many of the coping strategies (positive and negative) groups resort to in crisis contexts. 17
frequent exercises involving simulated WMD incidents can help inoculate against some of these stress effects and will allow members to develop proper coping mechanisms, learn from mistakes, and be less likely overwhelmed by the stress of a real-life situation.
(U) Another effect of stress upon the surrounding problem environment j I may be the problem of panic and mass sociogenic illness on the part of the public in response to a ROD or post-detonation situation (or one in which follow-on attacks are possible). There have been many situations in which an event occurs (i.e. Sarin attacks on Tokyo subway, Scuds landing on Israel during the First Gulf War) and medical facilities are nearly overwhelmed by people who imagine they have been exposed to harmful agents, even though they were not. For every one casualty actually caused by an event, as many as fifty other individuals may descend upon local medical facilities presenting with psycho~p~tic symptoms. Moreover, fear and panic may cause large populations to flee the 19~~tion of the event or the perceived target location, further complicating response efforts. :Thi~~~atters for II
du~ to the_ fact that the s~ounding con~ext is quite ~redic~ble!:~~-wili'~~~in
~
semor pobcy-makers. Pohcy-makers wtll be under mcreasmg:pres~w~ to talCe action of some kind, and will quite understandably become.,C!ven mot~:.~enuuiaiP'g.~f information
. This suggests the real needl
-l~i~~velop':sirong, active communication links to policy-makers, while allowing ~.~!.'?r analy-f!~rfo remain immersed in their analysis without being pulled away to an~~eiitil~!ions} Assuring that the c=Jtas a robust capability in these attribution envirolffiliiiis requl~~~it' to have a supporting communication/liaison ability that will allow !.!.!-9-~~n'?.ti~~:;~uring such highly charged, stressful contexts.
.'"-:::-;
-~;:::.
(U) The organizational structure and ope~~~~:P!lL~]1gnJ
-;~~~~=' I also should take into account the well-documented physiological.e(f~ts of sleep deprivation and working under high stress conditions. Occupational health prdf~~~j9pals, especially those who have experience working with groups functioni~g. under~:liigh stress conditions, should be consulted about work schedules, nutrit~Qpiii aspects of stress, and psychological support services.
_ .. -~:~:;:!? :
(U) In addition to provi~\!lg~f.ol~~rvig~s focusing on the health o~
I participants, efforts to reduce the i.nipict b'f;'s~~:jnd other cognitive mind traps should include training
'exercises and simulaiirins to edtititec=::Jarticipants on the various cognitive factors that could un4et6ihbei~ ~t'&i:Qution analysis. By using scenarios that highlight the dangers of anch9~g anl~!'ltstrate 'o~~rcognitive traps that may alter analyses, CJarticipants will be niote~'aware:of the problems and can learn to counter them as well.
.
:-===~=~ ::: ~~
.
## (U) Summary Recommendations Sec~Forn (U) Perils And Pitfalls Of Groups
(U)I
!will face the challenge of melding disparate elements together to forge an appropriate organizational culture for its mission. Issues of identity and norms, group dynamics, and social networks must be addressed to enable efficient effective performance in times of stress. Similar to the air we breathe, we are surrounded by the culture (or cultures)
we are part of, yet culture remains largely invisible to us and goes unnoticed until its norms are violated. This is simply "the way we do things," as well the underlying ralues. soctl norms, beliefs, and history of the groups and networks to which we belong.
as a group composed of members from a variety of organizations, professional dis(!iplj.pes, and traditions, will face the challenge of creating an appropriate culture for its mi~~~q~, and melding these disparate elements together effectively.
:;-::::;:::..
.
~~ ... 'I!-
.....
(U) The nature of the interactions within 1
~ and the social netw~it~:~1J.~~h ~~~~~::
underpin group and individual actions, alSo pose additionaUhreats j r:@~is section, we will capture key components of group culture, 'd~cs, ait~~~cial~etworks to enable 1
Ito navigate across this potential minefield. C~~ti_~g the'~gh(social environment for j
~s critical and will help establisl_l:~si:ro'ii'~;f.~un~ilon for future performance
.. ;:
-~;;:;'
.. -.-::::-:..
...~:!l~~::..
_;-'' .
(U) This article will illuminate a number o.f;pitfiill~;;ipcludiij"g.;t)l.t!Challenges of overcoming a clash of cultures to forge a common gr~~P:J.~ntitY.!:::This artl~f~'will also cover the importance of building robust social netW'&r~,acri>ss community boundaries, the dangers of
.. , ........
group think and other maladaptive group dyii~fS, and the natural limitations of experts.
Furthermore, it will also provide a number of ree:6b:\QJendations for overcoming these challenges.
..::!:
:~::''
... ::::!~'"
.:
-~ :;~:::~:::~: ::::
(U) Successful Culture in Jlig~~:stre~s, ffigh Ambiguity Environments
(U) Whilellfae:~ c~~i~~ge~.:iP integrating members from many different organizati~r~s:and c6~Qii1ties, it has the advantage of being a newly created group, unencum~r.~.,\yit\1:~~;.9wn historleal baggage. There is an opportunity to be successful and effecti':'~;.froili;~$? start;~~W.l~_!ng a culture robust enough to cope with its inherent challenges.
Crea,Wjg ash~ group ia:entity which members value and relate to, with a base of healthy, appib'pij~~ ~Btms for behavior and interaction enabled by inclusive social networks, will position"itl~~ll. 18
..
~ ,;. ..
(U) Reco~'l~donforOculture
(U) We reconiinend actively promoting the following elements to help navigate the minefield of group interaction, and regularly evaluating the group and its members.
-
Open communication and information sharing
-
Minimal in-group status distinctions
-
Inclusiveness (permeable boundaries for belonging, especially with regard to analytic
aspects)
-
Trust in intentions
-
Common group identity
-
Shared sense of goals and mission
-
Role of devil's advocate or red teaming
-
Understanding the strengths and limitations of expertise (link to expertise section)
-
Healthy interactive strategies, such as negotiation (link to negotiation section)
(U) Rather than conceiving itself as a team or group with specific members called in during a crisis, I
I can think of itself as the core, or hub, of a larger connected community of analysts, law enforcement, and more. Success in reaching this goal will require strong leadership that sets an example of cooperation, information sharing, and openness to new ideas and sources of information for the rest of the organization. These behaviors can be incentivized by making them part of the metric for professional success.
. .. .
(U) Group Dynamics and Malfuncti~Q'$:::
,
(U) Another factor is the pot~~ti~{!for ~~up dynamics or malfunctions to undercut ~
performance. During disc~i<?!l~ lnvQ!ving its structure and membership, 1;::.::_:__:_:._:__:_~;-::-~=:;il;;-l-=n-=ee:-jd to take into account how;-it ~ilf appro~h its analytic or investigative tasks, and how it will interact and commw:M~~ie wiii:i~Q~~r:~ctors. Under normal, non-crisis situations (i.e., low threat to sig@ti~ant::\(~Lqes, no time constraints, limited stress), groups have the luxury of explori~jfOpti9~ imd"g~PJ,~!i,ng information at a leisurely pace. Policy-makers will not be on the phdne eve..Y:;ien minlif~~expecting results from analysts, and tests will not have to be rusliea;::~<.:>r wjlfassessmeiits be made on the basis of incomplete or partial information. The normal f)~~ that groups go through in terms of developing working or personal relationship~~:;iQ.:g~tting to know one another so that disagreements can be expressed comfortably, ~~fin obtaining an adequate understanding of the lay of the land (regarding the
SOPs of the organizations involved, where information or expertise is housed, etc.), have time to be ironed out. It is unfortunate that in a crisis, groups- especially newly-fofl?led groups or ones that seldom interact- do not have the opportunity to establish these links.
This can lead to a substantial number of group malfunctions, problems of communication, and difficulties of interaction, not only within the groups themselves, but in communications with other groups as well.
(U) "Groupthink"
(U) A large body of academic research, including the well-known book by lrvinf Janis,
"Groupthink" 19 and the Hart, Sundelius, and Stem, eds., "Beyond Groupthink"2 , has
(U) Stllge of Development
..
.
.. .
(U) The stage of group development is a factor that has significant impact up4~:tfie performance of groups during crises. For example, Stem noted the problem of'N"~);v~Gro.up Syndrome during the Bay of Pigs for President Kennedy, and how the o~~~t.?,fthe"i'ii.~~~:after only a few months in office had not allowed the national security grouR;llie'~~ideni"reJied upon to adequately form or establish their footing. As a r~y]t, ther~(}Y,~ limi";~;pebate, largely uncritical acceptance of the intelligence and plannirig presented: bY. the OIA and JCS, and an overall group process that would lead to what Janis latei:4~_scrib~:~:a policy fiasco.
A more established advisory group surrounded Kennedy l.~;:yiori~:;J~ter .. 1t had developed working relationships and knew: the "lay of the land." It~as not as'v.\i}iierable to the same new group problems and handled the Cuban ~i~!-!~ Gti~~~W an exeriiplary fashion. For the c=J it is critical that the attribution com~t;niliity~i~ws the:grql,lp as one that is well established, familiar, unified, exerc.ises roJltinely, aiiil has str~ng communication systems in place. In the midst of a WMD crisis, if~ld~~-yitili that thes~ things are established beforehand, simply because there will be noi:im~.~o learn it on the fly.
'!':~.::.;;:.~.... . ...
e::~::~
'
~::: ~ ~
(U) Membership
-:~}=
(U) The issue of group membet'shlj;(~~~ the participants are within the group) is also potentially significant fo
.. rCJ.-!
,Lea~
__ ng aside the technical skills or expertise that members might require in the Va.9.P.psc.=Jgr9!J.Ps, there are basic issues of group dynamics to consider when these,ptayers CQll)e:t.()gether to address the WMD attribution problem. For example, i~ }.J,~.~ee~~;w.i,qely demo'iiStrated in social psychology literature that the status of .
individ~:~i)~~;pieniti~~~J?l!!-YS a major role in determining which participants will dominate the d!~.~ssion84;!ind whe"tlr~"Fthe group is able to think outside of the box on issues. Ideally, groupS\;~9.~14.-Iiot be top tieavy with only one or two high status individuals, since this tends to reduce::tJi~:.participation of lower status group members, especially if their contributions or views confli~;~).tpthe standard view held by the high status individuals. Also, if high status individuals are:;piirticipating in these groups, it is recommended that there be balance from different orgamzations/institutions in order to prevent premature closure of discussions.
(U) Multi-disciplinuy Advantllge
(U) In terms of the classic groupthink malfunctions (i.e., those arising from highly cohesive, insular groups under high stress/high stakes decision contexts); it is possible that the multidisciplinary/multi-organizational composition of I
I may reduce some of its vulnerability to these problems. By its very nature,i I will be a less insular group, although efforts should be undertaken weii ahead o tts acftvation to ensure the kind of communication and building of healthy cross-organizational relationships that would better integrate the more insular, individual communities that comprise I
I. What is important, however, is to not overreact to the fears of groupthink in ways that would
## Se~
undermine effective intergrrup crrdination in a crisis. For example, a phrase commonly heard during discussions of functions and organiZation is the idea of keeping the TNF
analysts isolated from the IC or law enforcement efforts in order to maintain their objectivity.
Yet given the overriding needs for information sharing and coordination that a real-life WMD
incident would demand, this would be counterproductive.
(U) Indeed, the challenges of coordination, information sharing, cooperation across these three communities cannot be overstated, and if not facilitated will result in serious inefficiencies, lack of coordination, and breakdowns of communications that could be seriously detrimental to the 0
mission. So for example, while on one level it might make sense to maintain separation to reduce biases being introduced into the TNF from the IC or law enforcement efforts, this potential problem would be outweighed by the probl~ms posed by lack of communication and coordination, and by the benefits that would b~i9.ti~ined by enhancing these elements. In addition, the multi-organizational composition I
I also should help to reduce the problems of anticipatory compliance -a group .~l:!lfunctl~~;:.i whereby group members try to please an external leader in advance by pi\T1iij:i{g what"''"
information or feedback they would like to hear, rather t~
~hat th~:~yidence::~\lpports.
=:=~
.: ::~t:::.. .
.
:~:
~::...
.... -.-~"-...
(U) Expertise: The Good, the Bad, and the Ugly
... ' :::;:;:;..
~:::.
(U) Also important to consider is the impact of perceive.4 ~~pertis~:~i:tliin groups of decision-
makers. As a rule, experts (or those group merol>.~~ pei'~iye to be experts) tend to have far
greater influence on group discu5sions, fnupiilifisW.~s an(i'i!!.e.rnatives, etc. than do non-
experts (or those who perceive themselx~~~~'? be le.slexpert)>!~s a result, group process can
become skewed towards the views of oni:Y=~:J:e.w .. eiperts, and potentially valuable dissenting
or alternative views from other group memb~~~ be self-censored. Within such groups, a
serious consequence of this dynamic is the shorti!~iJ"~}liting of more involved debates over
evidence, reduction in the variety of input _P,mvid~$ffor group discussions, and premature
closure of debate.
/:;=
.
(U) Fori
~.O~;.~:;oups operating within the IC, LE, and TNF communities,
it is important to asstgn.~_r~.team toAhallenge the expert assumptions and to encourage a
wide consideration ~f ~piiori~f:~!~-~!>;P'nor to closure. One solution to consider is the notion
of having a :~p_llec~9F:~:within the group whose primary role is to speak individually with
other m~nib~~!i9 eiici~~~ws_and perceptions which can later be presented to the group as
altema.tive hypqtheses. Ati:esipert opinion should be carefully labeled and considered, and
shouici:~:9eP.ciident upori"ihe scenario at hand. One's area of expertise may not align with
the actuiit:ev.~nt itself, since domains of expertise may be specific. The value added by
having spe'Ciii~:~xpertise, however, is the ability to rapidly synthesize information into
meaningful ch\!riiCs, recognize patterns in large sets of data, process information faster, and
identify incongruent data ..
(U) Studies have shown that small city or rural fire station commanders with twelve years of
experience were less expert than those with onl~ two years experience but who worked inner
cities with multiple four-alarm fires per month. 1 This point also relates to 'the status issue,
since participants with experience also may be perceived to have expertise by groups, when
these may well be different things.
(U). The AU-Star Problem
(U) A related problem for group dynamics and composition lc__ ___ _,lis the "All-Star"
problem. It has been observed in a number of professional level sporting events, that all-star teams - that is, teams created by joining the most exceptional players from across the league
- rarely produce the best team overall. While their members have exceptional skills and are tremendous athletes individually, these all-star teams typically do not perform as well as expected, or nor do individual all-stars perform as well as they performed on their originating team.
(U) I
lwe need to consider mechanisms for melding a genuine team out of a group of experts from different domains, with different organil'Jltional backgrounds, different cultures, different lexicons, and so on. Can we foster healthy, well-functioning social networks to enable efficient communication, work flow, and trust in a crisis? How should we assess success? This suggests that criteria for team membership should be a ~~~~e of expertise and the ability to work effectively with others. More often than not,::f;;p~itive group chemistry trumps individual dynamics. On the other hand, it may be possible;(<;:~. on the side of expertise, so long as there is a skilled leader who can manage the .P.<E~onal.iii~~!'.:<;:
,;~:,~-:;::~:~~~,
~-.
(U) Furthermore, for an all-star team, the needed types of C.:~l.?erts ar;;:95_fined ti~~e positions to be filled; pitcher, catcher, etc.
however, tile exacftyp~:of expertise needed cannot be fully predicted. The need for contributions from s~!}.~, higb:i*~pec;ialized experts may ebb and surge with the state of the problem. 1
I
I
.:::~L.:.....-r --..-;-;-:-\:;7;-;--,
- __
____,
(U) AU Expertise is not the same
.. :::::::=:::;~ti~: _;-:::::\~~=~::
| | -.. |
|-------|---------|
| .1' | |
| .. | |
(lD As teams ofexnerts are selected I
/
L
~t is important to create a mixture of differing types of expertise and should not be tii'6iJ.gg~.of only in terms of topical areas of specialty, but instead in terms of differing_.!y.pes o~~competencies .
.,P::'
,'"
.-.. -
(U) This recognizes that not ~l~P.erti~~ is the same. Some types of competency are more focused upon accomplishing=~~n ta_Sks, such as in TNR analysis of nuclear materials composition, whereas otl}.er\~~s of ~.Qnipetencies are broader in scope. One example would be the Director of a l!a#ii'na!' fa~At9'ry overseeing and directing teams of individual technical analysts.
:::::;::.
~ -
. : ~ ~! :! :;: . - . - :: :-,
(U)/ .. , : .. ::~1 it is 'i'iii~ttant that a mixture of expertise (task and general) is included witliiri:~n.pf i~:subgroups: Given that narrow, topical expertise may find itself out of sync with an'~iiibi~ous, new threat environment, the inclusion of those with general competencies will help s~i~~-~~
!flexibility to adapt to unique and unanticipated situations and preserve more'~$-iiOOng-outSade-of-the-box capabilities. Also, having broad competencies included with-those members with strong task comJetencies would add to I
I
coordinating and information sharing. capabilities.
## (U) Whlll Is Expertise?
(U) To be a genuine expert in a specific field, one must be much more than well-known and admired, or simply been around for a long time. An expert will be well-regarded by peers, able to render exceptionally accurate and reliable judgments in the domain, and will perform skillfully and economically while handling the atypical or unusual problems in their domain. 23 Some key points to remember are:
-
Expertise requires extensive practice, often a decade or more.
-
Expertise is domain-dependent (i.e. chess experts know chess, but may be mediocre at
backgammon).
-
Expertise requires deliberate practice. Time of service is not enough; experts focus on
specifically improving performance.
-
Experts see differently. Sophisticated mental models of the problem allow experts to
make distinctions, which others may not even be able to perceive.
-
Experts can get trapped. Very deep expertise can lead experts to become inflexible,
biased by their years of experience, and enamored of their intuition - ((U) Implicit vs.
Explicit Decision-Making).
-
Experts cannot go at it alone against multidisciplinary problems.24
(U} Experts can solve problems within their domain faster, and more accurately tban less experienced or proficient individuals. And they usually perform best at solving:J{pown (but difficult) problems in their domain. I
lwill undoubtedly be facing probl~~ .. no qne has actually faced before, so the tried-and-true methods may not be appropR~te.
~=:~::::;:
-~~ ~ .:-:-::..
.. ~
....
..~-
-
-
p .. .
(U) Experts are adaptive. They have the capacity to fluidly.;~nd cr~ti~~'y adjtist;tl:ieir
pro~l~m-so~ving strategies, ~enerate new ones, and find alieina:!J.~epers~!~~~s:' This will be
a cnttcal skill for the experts!_
l
::::::~..
::::
.. {==.
=:;i:~:::.:.- ,.
(U) Self-herding Cats?
.. : :::.
_}:;~~t..
.::
(U) The challenges of bringing together a_gfb~p-~f:~xpe~~=:i;};'tJen mixing experts and
generalists, have been observed many tiri)~~;in va.rj"Oils circu~iances (see the All-Star
Problem). This. is a task akin to herding ca~::::lp:'groups with distributed knowledge and
expertise, lower than average performance ctiitlibonly has been found. But there seem to be
factors that mediate this effect: the networks of~~i!ll"relationships among group members
can minimize the problem.
..::::= .
.:
. .. .
(U) In complex situationswhe"fctirl'forJ~~ion must be pooled across several domains,
decentralized networks feaqfumfhealtby working relationships among many members have
been proven to be mon::.;~ff~~iye th~J{centralized networks. Centralized networks often are
hierarchical structure$;:whereiiiQ~~:ii\dividuals tend to have just a few relationships which
end up all Aq~i.p.g,m~Q;!l_few prominent others.25
. .;- ".:' .. ~-~~::;!:~.
,=:=:=~-.
(U) n~P.pears 'iii~ advan~~~-~f decentralized, well-connected networks is that the flow of
kn6wi~8~.1!f:ldmformation is maximized. Lack of shared baseline knowledge or expertise is
overconie@.Qugh connections to others who know what is needed. What would otherwise
remain isolaf~:P.Qekets of knowledge can more easily be integrated. The existence and
functioning o(ili~se networks can be studied, and methods used in social network analysis
can better enable visualization and measurement of group structure and information flow.
(U) Recommendations for Handling Expertise Issues
-
(U) Select for fluid expertise in your experts. Encourage and teach fluid expertise
methods to all members of the group.
-
fiU-AFknowledge that expertise alone will not be sufficient to solve the problemsD
L__jwill face. Incentivize collaborative, instead of"all-star," behavior.
-
(U) Build robust, decentralized social networks for information sharing. Measure and
evaluate these social networks, and take steps to improve them as deficiencies are
observed. Share results with the group.
-
(U) Institutionalize these practices and metrics to shape the culture.
(U) Role of Negotiation
(U} In fulfilling its task, ~,---,Jwill be considering an array of ideas from these different communities that will be in competition at times. As noted previously, beliefs are possessions, often carefully guarded and reluctantly surrendered. Under situational pressures, including high emotion and stress, we tend to cling to what we know and believe, i.e., conclusions and decision algorithms that have held up in the past. The process of yielding a position or idea on a given subject is much like any other negotiation process. Two or more parties, all focused on the critical shared task of attribution analysis, will be taking different positions on a variety of questions. The resolution of those differences will rt;guite a giveand-take of ideas that will lead to a conclusion. As such, the concepts ofbowj~$1 rationality, bounded willpower, and bounded self-interest described earlier will apply.
' ;:!:::.,
-
. -~-~ .
-.;~:: .. ;.;,,:;
(U) Studies of the negotiation process have shown that participants in ~~:tt~~~~~ r~~~d positively, and behave more generously and are more willi.Qg to co~~g~r the po~i!ion of the other party when they feel they are being treated fairly. Mlitkers ot"faitri:re.atmerii can include being given an opportunity to participate fully in the decision-ri\~ing pro~~~s; attention and
~ct given to input, posi~on in ~e hie~chy, h?nesty, ~p ope~~~;s~ _.E~tablishme~t of an L_j culture that fosters th1s behav1or wdl result m a IJlQ: collaboraHYe and productive analytic and decision-making environment.
.. ::'.-.
,.::::::~=:.
--
.. ::~(:;~-:-::\)
::::~~~~~:~=
## (U) The Critical Role Of **Exercises** AnCjt~P.J-~Iiilding .
(U)j jshould meet regularly for joi~~=~~_xrcises to build relationships and the expenence necessary for a real crisis. Across tlie;:!mee communities, the key challenge will be to improve information sharing, cormnunicatioils, and build relationships to facilitate cooperation and sharing betwee~-~ey ~6imel, prior to a real-Iifec:==J"ctivation. By identifying seniorc=Jrepr~~en~J:ives"in advance, running exercises, and increasing awareness ofagency-specifiC' ~pabiliiies and information requirements, the group's performance in a real c~is ~ifl proy~;inore fluid than without any preparation. This is due to the fact that if 1
- Ji~ acfWAl~~quired for an incident, the normal learning curve of this important 8!'9-P.P. witi~J?.~ reduce~ ihereby improving performance. Having representatives from th~:IG::t:~.~~d ~.'?Oplmunities working closely together, sharing information in realtime,.-~9 serviQ:g' as comfi;ijibity-wide liaisons will speed analysis, enhance situational awar~ii'~~~~
'.
.
~ and reduce the risk of groupthink (which the communities might fac%~!;~~orkmg as mdependent entities). Analysts at NCTC who have access-to-the data system8;~f;9,~S these communities should support this effort. At different stages of an incident, this C9<lrdination and information sharing hopefully will allowt=ho perform far more efficiently at integrating the attribution process.
(U) Recognize the Need forD Inoculation of Personnel and Information Networks Prior to Events
(U) In considering how to structure II
and facilitate communications, it is important for C]personnel to stand up mor~tly than just during a crisis. There need to be ongoing table top exercises and drills, as well as team-building activities, that help to familiarize people with each other (and build the personal relationships and trust necessary to facilitate communications and information sharing), and also create awareness of the unique
challenges faced by I
~cross different types of scenarios. One advantage of running simulations and other exercises is the inoculating effect on the participants: This inoculation may include reducing the need to learn who other people are in their groups, or in outside groups with which they will have to work. It also builds personal relationships that help encourage information sharing and a greater ability to challenge assumptions or 'arguments raised in group settings (since people with pre-established personal relationships have a greater comfort level in challenging assumptions and arguments raised by people they know well, as opposed to relative strangers).
(U) It also will facilitate the development of people who could become liaisons across the three communities, due to their increased awareness of the issues and information needed by the different communities and their personal relations and comfort zones. Th~~ is,guite similar to how President Eisenhower organized the Policy Planning Board (P~~:).:~on his NSC.
He wanted his NSC policy planning staff to work on exercises and plans for scel:\~~s th3lt did not yet exist, primarily because this provided experience working thq~FY,pes of:~~s
(thereby reducing stress when the cases happened in real life) and beca~sifft~prQvide(l'
.
experience in interacting and challenging each others' assu!}!ptions. 4~i~enho~~:9:id not want his advisers trying to acclimate to a crisis context on the tl}i;ilor to.:i~t~:~?'perie~ce interacting and challenging each other (or him) during delibem,tiP,~ H~~~~ted that rperienle to ~xist prior to an eve~t. This underscores th~:~~lue(it.!~?c~tati?n of groups. 0
, and It also suggests that 1t would be best not tQ havec=Jw~eractmg only durmg stand ups of actual WMD events; team memb~~sJloutd,~~._interacting' continually.
,,/'':,:;~,\~1~: .
:~;~t:::.;/
(U)Estllblish Metrics to *Ensure* Effecti~~~~.o.[.lfcercises ~~~Training
(U) Exercises and training are essential to ~s'\i:~g healthy and effective group interactions and accurate attribution analyses. The effectivii~.ofthese activities must be measured objectively, however. As mentioned else"Y-pere, tiji~should include exercises under realistic, high stress conditions with active red t~1ng to ~hallenge the C=:J response capabilities.
In addition, the performance 9f:tb~~~group and individual members should be assessed by an unidentified observer so thafpar.tibipaqts do not behave differently around the observer
(avoiding the "Hawtho~~ E!~t"). ~~ermore, there should be a thorough debriefing of the successes and faUuf~:Of ea~JJ=~~rcise that will include an anonymous 360 degree evaluation .'?-f-~!f:ch P.~.~ipant. Finally, social network analysis should be conducted of email, telephoq~?an~;ffi~el-pe~Jll C!)mmunication among participating members to assess the exte~t;~d na~ of therr=~6rmation sharing and sourcing.
,..~=-~;~:;::a~
)
... ~
~~==~~~:-.... -~
## (U)Pres~Iif~~Plits.That Present Differences Of Opinion To Policy-Makers .
(U) Much as p~gident Eisenhower insisted that policy splits between departments or advisers be preserved in NSC papers produced by the Policy Planning Board, so that disagreements or uncertainties would not be papered over prior to his being able to consider the competing arguments, it is critically important that c=Jproducts also preserve splits in reporting for senior policy-makers. Although there is the understandable desire to provide a consensus response for policy-makers, this is offset by the magnitude of the decisions and actions that senior policy-makers will have to take in response to a highly charged WMD event. Even if it adds to the uncertainty and ambiguity of the policy context, those very elements could be important factors for their consideration and assist in avoiding anchoring to the usual suspects, and other cognitiv~ errors. Given the differing speeds at which the three communities likely ~ill enter evidence into the system, reporting these disagreements (or areas where there is an argument for withholding judgment until further analysis is completed) would greatly benefit policy-makers by providing additional context for interpreting the data.
(U) Counter Debilitllling Effects of High Stress Environment on AA G by Employing Stress Monitors
(U) Groups operating in highly emotional and stressful environments, are at risk of having their attributions biased by tl).e effects of hot cognition (the intense emotions of the moment) or by stress-based group malfunctions (i.e., the perception of short time, overreliance on stereotypes and other shortcuts, premature anchoring, etc.). Having group members educated about the effects of stress and emotion on their cognition will help them to serve as stress monitors for any c=Jgroup in which they serve by watching;f,?r J!.ebilitating symptoms in the group and warning against them during the attribution proces~~K
~:~~~:~!~!:;:.
;:
...... .. ..
.. !:;::;~:-~.~~
(U) Emph11Size Information Sharing *and* Enhanced~
Coordinatitili::;;~;;z.(,
.... ;.
(U) Given the overriding need for information sharing and . ~9ordin(\~~;: duri~~:~;;)VMD
incident, should focus on taking steps to avoid inl!'ffici~m5ies
1fU~c.ommiihication that could prove detrimental to theC]mission. This should be gixc:.l:! prioliij;~\ier concerns about groupthink. Any reduction in biases being introduc~(:t'~intO'~, f9r'example, from the IC or LE efforts is far outweighed by the problems posct~U?Y lack of~mmunication or coordination, and the benefits that would be oJ:!~ip,~ py:~}Wancing tfiese elements. As an organization composed of members from wiiltipl~ljgenci~~;:~9.1P-'Within and outside of GovernmentJ
I is not the kind ot:J.W1m'Y, ip~ar orgiji'i~tion susceptible to groupthink. nius, the focus should be'upt:)Ii:~.aiicing inforination sharing and coordination.
~":=~~=~:~ ...
## !'::~::;.. . (U) Summary Of Recommendations ~:R:
(U) Promote a healthy cultur~.,.,...... .,);;:?
/
(U)J
Jas a new org~izatj'on,.~an seize the opportunity to explicitly shape an orgamzatlonal culture desigije~to opt~ize functioning and mitigate or eliminate the effects of known negative factotS. This culttire could include behavioral norms and attitudes such as:
=~~~iy~ ._,
;:~::~;~;~;~~~~~t
-
Facili~~;p~n:c6tpro.unication and information sharing
-
Mi,Wfiiiie'iP.*iroup s~~-distinctions
-
.~!i~e in~iiisivenesfi;(iiermeable boundaries for belonging, especially with regard to
aniil~ic.aspects)
... ,. .....
-
Trust iti:~~~?~9ns
.
-
Create co~on group identity
-
Establish a"~hared sense of goals and mission
-
Incorporate a role of "devil's advocate" or red teaming
-
Understand the strengths and limitations of expertise (link to expertise section)
-
Practice healthy interactive strategies, such as negotiation (link to negotiation section)
(U). Maintain and grow social networks
(U) Among their many benefits, well-connected and well-structured social netw~rks improve performance in groups facing complex problems. c=J members must maintain and leverage their existing networks while building new relationships both within and across group boundaries. Regular, formal assessments on network structure and individual position and performance within the network, using methods from social network analysis, will provide a
means to evaluate the current functional status and suggest areas for improvement.
(U) Manage Expertise Effectively
(U) Expertise, while often critical, can be a double-edged sword in groups, generating conflict, rigidity, and excessive deference. To leverage expertise effectively, I
~hould
consider:
-
Selecting for fluid expertise in its members, and encouraging and teaching fluid expertise
methods to the group.
-
incentivizing collaborative, instead of all-star, behavior. Expertise alone is not sufficient.
Metrics for individual success should include cooperative behavior and information
sharing. Provide actionable feedback to individuals, such as a personal social network .
analysis.
:: .. -:~:
-
Building robust, decentralized social networks for information sharing 1,--.::.::.w"- ---..,
c=J Measure and e~aluate these organizational social networks, and take si~p~,_to .~
improve them as deficiencies are observed. Share results.
,.;;~:~:;:;..
:;~::~;::-
-
Institutionalizing these practices and metrics to shape the culture.:;-.{
;~'~;;:;:~: . .
.
:~~!;
>=~ ~~~;:~: .
~~ :f~~~
(U) Mitigate maladaptive group dynamics.
k:-.
-:;;;:;;,_ ...
- ~;;~!;~'=; ~
..~;. ':;""
-
(U) Avoid over worry about groupthink, but also avoiq.j!J.sulat.ji:y,,ofpei'spectives.
Consider membership criteria to create a balance witbiii"the gn)iip~{aeross status,
expertise, etc.) and recognize problems ~~S?f-~~ed,~iiji:~ewly forfued groups.
-
(U) Recognize the need for inoculatiol}.:ofpers~!ptl ana:i:9f-9~tion networks prior to
events and meet regularly for joint ~~~f.sises ro.:J;uild the ~hitionships/experience
necessary before a real crisis.
-:~;:;::. _/,'
-
(U) Preserve splits involving differencesi8fiqpinion in reporting to consumers and
emphasize information sharing and enhancMt:=J coordina ..
ti_on_._--,
-
(U) Counter debilitating effects of a ~j~ stre~s-~environmen~
I by employing stress
monitors.
. . ..
..{::
-
~elect a mixture of e.?C-Pe~:Witli' differing competencies (both task and general) for
L__j subgroups to enh~cc:.:flexibijity in dealing with ambiguous environments and
coordination across_t]le ~communities .
.
:)~~~
~:~:::::::~:;:::;-~
## (U) Bringing The Right Broomstick To The Wizard
(U) This article focuses upon key elements of the senior policy-maker environment and how these may pose challenges I
lin effectively communicating its attribution assessments.
These elements include: how senior consumers may view threats and risk, differences in accessibility to inner circles, the right kinds of experts to communicate with them, and the importance of making I
Ia key information hub for decision-makers.
(U) Understanding the Sensitivities and Constraints of Senior Policy-Makers
(U) One of the key challenges!
~ill be to not only produce the highest quality attribution assessments, but also to understand that the consumers of its informati.Qn (senior policy-makers) will be deal in with olitical issues and perspectives that may:ij~ye significant impact upon ow its assessments are used. I
1.
I
lit must unders at not a 1 policy-makers use information and:advici;ih~~thc;
same ways. To maintain effective communication or input into the polj~ypf~_~S, it"fs'
essential j I to understand hQw policy-makers di!f~r frolll:iP.9~ .. anothei-;jipd how they d
. . .
.
....
.;
.....
'to'
respon to cnsts sttuabons.
.
~;~~:;;~
.. ,
.-~~=j~.
.
~:.::~~~:y~
(U) We also must consider the state-of-mind of the decisio11-~.ii~J5(~ well~ the analyst providing data and assessments to the decision-maker) .~'!.~~~uent io:~;;lilgh consequence event such as a nuclear detonation. They wilL~::11.g!tated;;~9rried atiout family and friends, and may well be in a vengeful state-of-m~9~~iti:~s9 coni~k:~~z-~ould a decision-~aker be willing to react with punitive actions, ey.~il:oifthe c~inty ofPl'e attributions or clarity of the evidence may not be there? Will their'sti.fi~'J::.otproof change? Will the psychology associated with the event and the pressures fo~:~~tion reduce the decision-maker's standard of proof and his or her demanded level of certaincy:t~;QJ.e'assessment process? We need to also consider the public demand for the govei'I}P[lent to:;tiike action in the aftermath of a catastrophic event. This can cause tretp.~iiaous pressure to shorten timelines, and lower the threshold for action. It could.b"e:JjiUdeiii to define certainty thresholds in advance of an event to help circumvent these ph~o,:nena. :.
!~ .. ::"'
. ::
(U
it is important to undersla!l9:,~!=ver.ali~~~ues regaroing the sensitivities and constraints that are likely to apply to seni?f.:-P9li6~f:Fers::~~P.~,the activation ofj j
.. ~;::~;~=:~.
)!:-
~~=~~~
(DJ Th~t::JAnalyst as Consultant
(U>I
dependent on a host of other factors and constraints. Information must be actively incorporated into the mind. New information has to be assimilated into a person's preexisting
## Secre~
context, or state of information about the world. "Meaning construction" takes place when this new information can connect with what is already understood. Our ability to absorb new information is limited. People selectively attend to new information that connects, and may be oblivious to the rest.27 Emotional states, such as stress and anxiety, as well as the degree of cognitive openness, can limit one's ability to wrestle with new information, especially if it conflicts with previously held beliefs.
(U) In the non-routine, hard problem world!
I, information seeking will be part of a process. The early stages of information seeking are likely to show a marked lack of clarity and precision because the specifics of the problem are still imperfectly understood.
Confusion, uncertainty, doubt and frustration are natural, and a desire to just drop it and move on to some other aspect of the problem needs to be tempered with realistic expectations of how information seeking actually works.
;!~;;{''
.. ~;:::;:;:~ ~
.
(U) Later, as new information gathering starts and understanding of the P.IY.~!em ifiiP.~9~bs, one may experience relief and satisfaction that the needed information ~h:g'~~J? . suppllea. Conversely, one may experience disappointment that no re~Qlution ~~r{be aclil~v~ even though the information need is clear and focused.
:::-;
!; '~;::~:..
?:
.
.
..... _
~ .
.,..
.
(U) Paradoxically, only after you know your answer are y~~=~~~~~;Y?-.kn~~;t~ur question.
3:~,~~- .-
-~~~:~~;:~~
(U) Performing a Reality Check on Information.;Seekiri'g:;g;.,
,;
(U) From the perspective ofthe analyst
.. ,,,.,,.,,,...
,
I
I there is a compleme~~-'-:
-
.,...,,.;:p-~--cob:-:1:-em".
;-7\. --=Th=-e-p-ro""=)j"'=le_m_ma_y_tr_a_n_sn-=-o-rm---:fr=-o-m----::''what do I need to know to take action" into ;wlia(~(!<;,>.~s;my client really need me to provide?" By taking the perspective of analyst as consultantl:J:t;~ill require more than simply answering the questions that were asked. Optimal performance:l~:!aformation seeking in this context has to be oriented around knowing the mind oftheclieni:::
.. ~;~'~
.
-
(U) What is the goal ~r:fii~;!;c~~der's intent?" Questions may be tailored to
address what they p~cejVe cari:be answered, instead of stating their actual goal.
-
(U) What is the.;qtfxr:~tion Pr::<iecision to be taken?
'.: - .: ,..
.. .lo ... :!i .... - .. ~~;;~:-~
- (U) What lev.~tof detaihrs:fiiily needed?
-
(l!.):J.~;:!8~re:;afi~~~r authority/group beyond the current client to whom this
j.nfonrl~~pn nee~~g.,be presented?
(U):Q~j"
=r:ctivities !;n undoubtedly go beyond providing a single final assessment report. T&:~~~le ~ew information needs in a dynamic environment, the process must be interactive ana~;~~lf-corre~ting to whatever degree possible, not "fire.:and-forget." To confirm understanding,<5fwhat is actually needed, llmay choose to implement steps such as:
-
(U) Provide interim feedback on th'estateof the work, and allow for course
corrections.
-
(U) Provide samples of answers -they will know it when they see it.
-
(U) Rule out the irrelevant or superfluous with counter-examples.
(U) Is it Safe to not Know?
(U) Acknowledging ignorance, that is, lack of specific knowledge of expertise in a social environment, can be a risky proposition. Given the complexities of the c:=Jmission and operating environment, and its composition, both lack of joint expertise and of situation-
## S~
specific pieces of information are a given. I
!will need to create a social norm that reaching out to gather information and expertise is not just appropriate, it is required.
Acknowledging lack of perfect expertise, and taking action to rectify the situation, must be valued normative behavior.
(U) Recommendations Relating to Asking the Right Question
(U) Dealing with the social and psychological factors around information seeking will be critical!
~ The following should be considered as potential methods to improve fllriction:
(U) Providing education on the information seeking process, to avoid common pitfalls and overcome natural frustrations, and be able to optimize their own behaviors. ::..
(U) Creating social norms that accept imperfect knowledge among members, .~4;{eward
outreach to acquire the right knowledge and expertise.
'
::;~;:;..
(U) Tapping into the mind of the client.
.:;:=::::1::::-.
:::~~;(
:;:~:= .
.l:o" ~~~-:~ ..
(U) Recognize that Pressure for immediate (Short-Term) Political or Pfl_ljt;y R;J~~~f!; WiU
Be Directly Proportional to the Magnitude of the Consequences Surniiiidi~g an E~int
/;. (
~~~;~~~~:.~.
..
(U) In the event of a substantial WMD incident, policy-maii~rs yviU:fJ~i~,~~me~~~~s pressure
to respond publicly to an event (especially to a highly visible w,Mp emplQ'Y;m'imt resulting in
substantial loss of life or contamination) by quickly asce~~g~''bJ~e o't'responsibility for
the incident, implementing strategies for managing the H~~onse, and:).~unching punitive
actions against the perpetrator(s).
. ...... ..::::}:,.
-
(U) Ahead of any WMD event,r--l~h~u,ld f~~~ upo~-~~h~~ting senior consumers about the reality that technical ~rliij~
.. ~"ene investigations, or IC efforts are likely to lag behind their desire for more rapid delivefY.:~f.actionable information. This is where managing the expectations of policy-makers thtQqgb.pre-incident education is important so that realistic understandings oftechnical ~<!Flbilitie~, how fast certain analyses are likely to take, etc. can be provided prior to the op,S:et of a crisis situation (when tolerance for that type of learning will be extremely,~i.nijt~d).':Pre-event exercises (ideally including senior customers) would educate tl!e"~.!c;-8ar4ing the reality of how fast they will receive actionable information.
.c :=:-(
.. ;::
~==~:~'"
~~=:~!-:;~;-::;::
(U) Recogn~~i/}!it:J;~kCJ!;-Ma::;;olerance of Risk (and wiUingness to IICcept various meiiS.f!ifis or es'j/!nates Of.?,e.iiamty) WiU Diminish as the Magnitude of Potential Co.nseij"f!;~:nces:'(On *EithefAction* or Jn~~etion) lncreiiSes
";:::::: .... ~:
(U) Reg~~i~~~:9f.how many different standards or measures I
lmight use to communicate ~j.iying levels of certainty to policy-makers about its findings, the actual consequencesof an event (either real or potential) will cause a subjective shift on the part of policy-makers in how they will interpret this information. For example, faced with the chance that a nuclear device might be detonated in a major US city -and a 75% certainty expressed by the IC or law enforcement that the device could be seized and rendered safe from terrorists before it could be detonated- it is likely that policy-makers would consider a
25% chance of a nuclear explosion in an American city too large a risk to take at that moment.
(U) During the Cuban Missile Crisis, US Air Force and IC estimates that 90% of 'known'
nuclear missiles could be struck with air power (while only about 67% of the missiles likely to be on the island had been found) provoked Defense Secretary McNamara to argue in the
## Se~ (B) (1)
ExComm that even one missile launched at an American city was too great a risk for any President to take (and a consequence beyond acceptable imagining). When consequences for an action are high (in terms of physical, economic, military, or political blow back), the standard of certainty policy-makers are likely to demand will be greater than under other consequences. In contrast, if such a certainty level were expressed about the possible origins of the material found in an unexploded, interdicted ROD (for instance, from stolen medical equipment in the FSU), policy-makers might well view this as acceptable proof for taking actions like contacting the. Russian government diplomatically to request information and assistance.
(~) The attribution question may involve a spectrum ranging from successful interdiction of nuclear materials that failed to detonate (which allows greater time for inve~tiga,tions, more careful analysis, and a process that takes place out of the public view) to thos~itqv'otving significant, visible consequences (use of a Radiological Dispersal Device, alsci"kli()~ as. an ROD, or nuclear detonation). In those cases where substantial military r.~.~M1ltio~~~q;#t adverse impact on US interests might follow as a result), policy-make~:M'i'li~~ly to iitSist upon a very high level of attribution certainty.
...
.~::::J
,;;~::::;-
:=:~~
;.e~~~~=-~~:
~~~ .. -
~)
!fte hypoth~ti~l ex~ple_ of a l~ely conne~tion to te~~r.jst ~~-~~~o;;h K~~ean matenalm an ROD tnctdent m Chicago Illustrates th1s pro~l~in. '!l~ause.any US m1htary strike against Pyongyang could potentially trigger a rea~.tibfi from itie;~orth (such as a new war on the peninsula or retaliation against Seq!lk9.x:.otl)~f;6lgional aUit!s), the probable consequences of the policy actions if the a~i56iii>~:js ace~pioo would obviously temper how policy-makers would view the degree o{,~nty'fli~y wouia:~re:<iuire regarding the evidence
. .
'
.
~ ~,...
~-
....
pnor to taking act1on.
.
-:~:;;,_
/
=:~~~~~~;;_
(U) Policy-makers facing substantial consequei:iC.~~f-9r taking (or not taking) certain actions will be far more risk averse (and require f~P.:~P.ighe~;ii:vels of confidence that the analysis is sound) than will those facing more limi_~;&r less'significant costs. In those cases, lower levels of certainty may be ac~p.~b.Je, ff'ihe costs of such acceptance appear not to be prohibitive. As cqiiS"i4~ v~ous measures of certainty for attribution, it should
take into account this b~jc ~~!illty: ~~!risk acceptance on the part of policy-makers i~
closely bound to thei!' percepij~~;~9ritlle consequences. It is part of the complex politicalrsychol.of;!~~~~~r~~~~;~.,w~ll "diive how policy-makers actually interpret the analysis from
(U)-~~;jJ.lhere~f.imprec~:i~:ofterminology for expressing levels of certainty together with the varY'iii.g:9-~grees of meaning individuals can assign to the same term suggests that making an effort to:~~bli&h a shared understanding of terms relating to certainty is imperative. For example, diffegnfindividuals show surprising contrasts in their interpretation of terms such as possibly, prbbably, likely, unlikely, and certainly. But even if the communities adopted legal terms such as beyond a reasonable doubt, clear and convincing evidence, and the preponderance of the evidence that have the appeal of familiarity to the listener, they still
.
retain the same kinds of imprecision due to each individual's own subjective idea of what the terms mean. This is an area where ~
lwill have to move purposefully to induce the three communities to work more closely together to reduce the risk of miscommunication by establishing a common lexicon and definitions regarding levels of certainty.
## (Ul Understand That Peoole Resoond To Estimates Of Risk Based Upon Whether These Are Presented In The Domain Of Gains Or Loses
(U) Prospect theory notes that people are risk accepting when information (such as reports or data being presented to decision-makers) is couched in language emphasizing their loss of current possessions, status, etc., but risk averse when such information is framed in terms of gains that also requires risking current security or possessions. Thus, /
~enior Level must exercise care in the labeling of various kinds of attributions, estimates of a current situation, or the chances of an operation being successful or not - since senior leaders will likely respond in prospect theory ways to the presented information.
## (U) Not Au Senior Policy-Makers Are The Same
(U) One of the major challenges I
I in providing information t9;~~9::;:~~i~~:f::-:;:
makers (including the President) is the reality that no single approach fj.ts all:P.<?l.icy-ni3.kers.
Regardless of how~
~tructures itself, sets up acces~.;~hannel,sJ~~:.senioi":l~~18, or packages its assessments, these things will ultimately run tip''ag~inst tH~:P,~rsonat styles and characteristics of the policy-makers themselves. Like all indiv,i$J;~ls, seiif~~policy-makers will vary greatly from one another in their needs for info~tior{g~t;!l~ring'""and debate (prior to making decisions), how much they feel the need to reJy:lij)on ex~eij~opinions prior to reaching decisions, and in their overall sensitiy;i~;~Q t~t;;:~~ounding 'policy environment .
... ;~;/: ~:::::~::
.
~:::~~:;:.-~
(U) This is reflected in the broad leadet:~~.P,..literat!J!:e that ha8;:Qeveloped over the years in the fields of political psychology and Presidd1q~~:~Wdies. It chrOnicles the wide range of differences in how modem American Presideii~:J!ls well as foreign leaders) have structured and actually 'used' their advisory systems durilig:"Cti_s_is decision-making as a result of their own, unique individual differences (i.e., ~r.;onalii};:bharacteristics, style, prior background, etc.) Illustrating this point is the book, ;':f}ie President and His Inner Circle: Leadership Style and the Advisory Process in f:oi~ign Affairs," by Thomas Preston which profiles modem Presidents from Harry Trull}in !ht'ougl! George H. W. Bush across foreign policy decisionmaking cases. 28 Here . t!t~ ijl.9ividuat~haracteristics ofthese Presidents played a predictive role in shaping how ~h~y:woill~:~~~fiire and utilize their advisory systems during future crises.
.-:~::::.
-
~-::;:~:;:::::::~ .
-::::~i ...
(U) D.Uferent ~s of ~~~id~~ts favored more or less open advisory systems, sought out more:a~:J~~s qiVerse sets df advisers, had widely varying needs for-information or diversity of advice, liiia;Dad differing needs for (or reliance upon) expert advice. In simple terms, this distinction~f~:J?-~.~ought of as a difference between leaders in terms of I) the degree of control or per8c?nill involvement they required; and 2) their own personal need for information and sensitivity to the broader policy environment around them.
(U) Leaders who are more hands-on tend to delegate less to subordinates and want more personal engagement throughout the policy or decision process. Leaders having substantial policy experience or expertise in a given area would also be expected to be more engaged and to delegate less to experts. Presidents Eisenhower and GHW Bush, for example, were much more highly engaged throughout the policy process (and much less dependent upon the views of expert advisers in coming to their own judgments) in the foreign policy arena than they were in the domestic policy arena (where both had less experience and interest). In contrast, those with less personal experience or expertise in an area will be much more inclined to delegate to subordinates and will be more dependent upon expert advice in reaching conclusions. So for Presidents Truman or George W. Bush, who both lacked extensive foreign policy experience or knowledge, expert advisers around them had much more of an impact (and were delegated more of a role in policy formulation) than were advisers to President Eisenhower.
(U)
f one is dealing with an engaged, hands-on consumer
(perhaps one who has substantial foreign policy experience), the attributions or analyses provided senior leaders may be used far more selectively than would be the case if the consumer were of the opposite type. Moreover, it is very important to have educated the more hands-on or experienced leaders in advance of crises regarding I
S
abilities and value in attribution assessments, so that a pre-established relationship will be developed that thoroughly engages with that leader's desire for involvement. :.
..
=t~:.:=J
(U) Similarly, less sensitive leaders, who see the world in more absolute, black~~9;:~hite terms, tend to be more ideological, rely more upon simple stereotypes or,~!ogies:t~;i:i~
understand their problem environment (or the options presented to thetiJX"iliii:{~t~d ti>"ll'ave more closed advisory systems populated by advisers who sb~re simiJ~iJ>elieti~~:views as the leader (and have less diversity of view as a result). Histb'ric exailip\~s.-~f such American Presidentswould include PresidentS Truman, Johnson, Reagaq;~:~p G.'W.:!~.gsh. More sensitive leaders (who show more differentiation in their ~~~jro~~JJ-! seeing the world in shades of gray and many perspectives instead of in absQ,lp~es) tend "titb.ave more open advisory systems, gather information from m1!.~~;~~r~~:d~~~r.se sources (that both supports and contradicts their own views), and pop':l~atethejt.~~visoi:t~Y.~tems with a diverse group of advisers who do not share common vie~~:9.r positipiis. HistciBe examples here would include Presidents Eisenhower, Kennedy, GHW":B4J:l,_<;::,liiiton, and Obama.
.. :~~-:-:
(U) This distinction between sensitive and le;~:~~~j~ive styles is important for I
I if the consumer is a more sensitive type of l~per, I .
I will be operating in an environment more conducive for communicating inf9tJiiation to senior levels (since these styles are more active in gathering informatio9.f~~jce from a broad array of sources), have more open advisory systems, and compfise .a.Ctors;.who monitor their surrounding environments. As a result, the contraction o,f..~u$.~rity pr<J:~lem will be less severe for I
land the problems of access to senior lc::v~ls:'les~::~.~~RU~ted than it will be when providing assessments and attributions ~'?.less ~~A~itive leaders: For these less sensitive leaders, I
1 should anticipat~;:iB!i@!l mo~:~losed advisory and advice system and greater difficulty in coiilll?:~icating!~ffectiveiY.,:}o"senior levels.
must recognize that greater barriers will~'i:i~t;.to gaiiilng the attention of policy-makers. For example, because less sensitive leader,; te;~p be more ideological, more likely to adopt stereotypes of opponents or analogies dii:Y,~Q by. personal experiences, and surround themselves with similar types of
advisers in te~:;ofbeliefs, it will be very difficult to successfully communicate that the usual suspectsfuight not be behind a given WMD incident, strong, pre-existing belief structures notwithstanding.
(U) Generally, discrepant information has a great deal of difficulty penetrating to senior policy-makers in such contexts and, even if it does, has the problem of gaining the attention of policy-makers (or perhaps more importantly) the perception of 'credibili 'for them if it conflicts with pre-existing views. This is why having an informed, active ctor as part of the inner circle for such leaders (who then provides credibility to assessments due to their own status within the leadership group) becomes so critical. This relates ~o the idea noted earlier that beliefs are possessions and that a major hurdle in getting policy-makers to think outside of the box in attribution environments requires a willingness to challenge preconceived notions or views./
~his will become most problematic when the consumers have less sensitive styles because those prior beliefs or modes of operation that previously were employed will represent a simplifying heuristic (or shortcut) to them. It will require more effort for an c=:Jlttribution assessment that violates these to receive a fair hearing. More sensitive leader styles will be more amenable to considering these outside the box assessments and will not rely so heavily upon preconceived notions in framing or
-understanding the policy environment or situation.
(U} Bottom-line: If/
~s dealing with a President favoring a more open advisory process, diverse sets of advisors and extensive collection of advice and information - the task of obtaining access for providing attribution information will be far easier than it will be if advising the reverse style of leadership. Although prior education of the existing White House regarding capabilities, managing expectations, and setting up relations~ig{for c=J
technical advisers should be done in any event, with less sensitive leader styles"iil"d.more closed advisory systems, this becomes critical to avoid serious communi'?{l~~n p;6t;!~m~: in the event of a crisis. On the positive side, leaders with less expertise in:iif~~-.will Se'inore dependent upon expert advice or explanations of the problero envirqp;qient, sci:~j~hproper access, the analytic product I
I could be of great ~~istanc~iiftib,J,se for;~uch leaders in understanding their situation and evaluating options.
/:::..
:::;~;::;:'
{U) c=Must Calibrate Communications with Senior c6g~m:!~~1ea -~pon the Way Different Leaders Structure Their Advisory Systems /::;:;;:;..
(U) Because the flows of access to sen!.o~i!~~_:_:~_ ';l..,ircl:~=~;;t oolirmakers (such as the President) change from one White'Hoilse;Qc.cbpant to the nex must take into account differences across consumers in tenliS:~~f:J:;10w ~ey use advice, how much information they tend to gather, and how attentiV~:!ste'ihey to their environment in setting up their communication strategies.
..;:)
.. :-
_ .. : .. :-"
.... -
.~~;~~
.
(U) Leaders with closed adv\soi}i::s"ystems and tendencies toward limited (highly selective)
information search will proy,'e ~p.O~e difpcult for C:=J to communicate with effectively than will consumers with m9~ o~~ sys~~Qis who actively seek out broader ranges of information.
M As a result. c--j:r:t e~~~i~~{~ommunication conduits to senior policy-makers
/
~ These conduits must be sufficiently robust to pe~?.9Y.~ffec~~~ly ~d ~aintain access in cases where leaders have relatively closed inner circles~Q'~!~ll~ the contraction of authority during a crisis makes even open systems more imperm~il~J.~? ....
:::~:::::~;~ ..
## (U) *The* Con;~~Tion Of Authority Problem
(U) The Contraction of Authority problem reflects the reality that the inner circle around the national leadership tends to shrink during intense, high stakes crises to much smaller groups composed primarily of the closest personal advisers to the President. This occurs because of
1) time constraints, secrecy requirements, and the need for rapid decision-making often imposed by such crises; and (2) the desire ofleaders for a comfort zone composed of their most trusted advisers during such times.
(U)
this contraction of authority could potentially pose serious problems for its ability to effectively communicate its attribution assessments and maintain the kind of ongoing contact necessary to keep policy-makers adequately informed of new developments
## Sec~Orn
as evidence continues to be gathered and evaluated during a crisis. Therefore, prior to the actual activation I
!during a WMD crisis, it is critical to work to create and maintain an access channel to senior policy-makers that can survive ~ction of authority during an event. This could be pursued by: 1) setting up theL__Prganizational structure to include a senior Presidential adviser (someone who would be expected to remain within even the tightest crisis inner circle or have easy access to it), who would become an active c=::Jparticipant, take part in its exercises, and become a knowledgeable conduit for information to senior levels; 2) educating senior policy-makers at the highest levels about the c=:Jmd the functions/information it can provide so that it is instantly thought of by these leaders (even during a time-constrained crisis) as the main resource for understanding the difficult questions they will face and coordinating the views across the IC, LE, and TNF
communities.
~...
~ ..
;~:~~~"'
:-..~::~ ..
(U) Using the Right Kind of Expert Advisers to Interact With *Policy-m(!lce.:S* . -;:;::;:~ .,,;:
l'L>o'
"''" f.
(U) Another challenge is fmding the right kind of messeng~f"'of.;~p.ndui(iiot only for communicating the basic attribution information t~.senior pi)l~~y-mal(~i];:out also for facilitating the ongoing back-and-forth interaction required!
jtO,~~sist iri'coordinating the IC, Law Enforcement, and TNF communities in ongoing a@.f2:1:ltion taS~s:' It requires that the ~ght type o~ ex~ert advi~e~s) is still involv~d w~thin ~!f.t.innt;(~~~~le w~o can_ convey the avatlable techntcal mformatton I m a ttm~ly~and effecft~fashton wtthout distortion. Assuring the presence ofsuch an ~~~!1;W9u(4;~1Jow policymakers the best handle on the moment pos~;ble "(8$~;-W,ell aS'~l!ge the expectations of policy-makers), given the available attripjju~m infdffuation sO:jllat this sometimes technical data can be given its proper context in-'or<i~t:tq. afoSi.st the White House to improve its situational awareness. This would require eei~q~tjon of policy-makers prior to an event apout the value of including such an adviser, and a s~li:(>:f:p:rotocols for identifying this individual in advance so that there is no discmmect be~~en pojicy-makers and the best the technical community I
I have to offer dutirfg:;im intense, fast-moving crisis .
.. ~:; .. ::;::::~> :::.r
(U) Given that many senior .pol~~~rna.!cers lack technical backgrounds (having business or political backgrounds i~~t~4)~it is imPortant o be "bilingual," in the sense that they ar~ ,~iinv~~~;~d.imderstand the politicaVpolicy environment (and problems) faF~.d by;(ljc;;.White House, as well as the technical, scientific side of the attribution equation,fo~~C:f:.tiponi I. This would greatly facilitate a bridging of the gap betw~!~j:{those"W.Iiose foc~;(ilnd comfort zones) are in the policy realm and those who come from'ii~:i!t~re ~~i~ntific or technical community. Clearly, in a nuclear (or other WMD)
attributlciii!setting, it is of critical importance to reduce the chances of miscommunication or any lack orP.~~!-~i.9n in the communication or understanding 1
!data by policy-makers.
~-:::~=-.
.
(U) *The Role* of Blam~ce Strategies and How These May Complicate the Attribution QuestionL_J
(U)
Whether it be the IC, the TNF, or law enforcement communities, or the senior policy-makers themselves - during a WMD
incident, there will be a sensitivity and awareness regarding the potential political, institutional, or career damage that would result from attribution failures (for example, technical mistakes, incorrect interpretations of intelligence, or failure to pursue certain leads)
and a desire to avoid divert -or even direct -blame were they to occur. Similarly, policymakers themselves will be highly sensitive to the political damage that would result from policy mistakes (for example, mistaken attributions of blame resulting from retaliating against the wrong party, negative reaction to policy actions (even if the correct perpetrator was identified), failure to take action in the face of evidence later judged to be sufficient.
:: . . /
(U) The higher the potential or real consequences of the crisis incident, the greater the resultant emphasis upon blame avoidance strategies becomes -which in extreme circumstances could severely hamper trust and communication between the three communities An excellent illustration of this area of research in crisis management that focuses not only upon crisis response, but also the management of response and its aftermath from a blame avoidance standpoint (both for institutions and policy-makers) is found in the book "Crisis and After:
The Politics oflnvestigation, Accountability, and Learning."
29
(U) Because of blame avoidance, policy-makers (or institutions j
/)~;~~~!_fmq themselves seeing to selectively process information during the attributig!M~!ldeavorj~f,.9r example, assume hypothetically that policy-makers had a desire to adv#ce'~:Rortanrj:}eac~
talks or maintain cooperation with another country in a polj~y area Qfgfeat irii';;9rtance to the United States, and an attribution assessment arose that potei1tially lihki6~1-:that nation to a domestic WMD incident. Depending on the severity of the incJ~~pt, on~:%.9y.ld imagine in certain contexts policy-makers desiring to search the avail~~>Je' errQf.:pars iir'the attribution evidence to introduce doubt (much like a competent def~rise attorneyJ.~:,'fhis introduces
'politics' into the way in which attribution ev~~V~~. is . \ii~}Y.~d by poll~y-makers, and certainly skews it in directions unrelated tC?,:th~'C:]foc~~P.~ ~\J.e IC, legal, or technical analysis.
. ... :{h:..
.~;t .
}~;:
(U) Similarly, ifa particularly horrendous WMP., attack occurred and there was a lack of clear evidence linking a perpetrator to the crime, there:~C?gld be immense pressure on policymakers to avoid being blamed by the publi~;.for d().iifg nothing in response to the crime and to
"round up the usual suspects," so as to g!'Yii" the public the perception of decisive, strong action. This gambit could res9.lt:;w;!poli"<:y-makers taking actions against parties based upon I
!attributions that h~d vt;cy lovy levels of certainty by selectively using bits of the evidence to point the finger:in-tbe desi:fed direction. In fact, there are almost endless scenarios for how th!? ijiiioe;~Y,~iPM~~ game might be played or what confounding effects it could have,_~g~~ ~~CJ.ttriotifion effo~. ~is observation is not me~t to denigrate any party, b~t;me~IY.. to recl)gntze that the attributiOn process wtll not be conducted
.... ~
"- ... .
...
.. .
. ...
in a v~:ffilm, an~that pcili~~s- sometimes not directly related to the issue at hand- might playa=:ffit~ in btiw policy-iiiakers actually will interpret or make use of any attribution data I
I
~:-::~.
(U) Because~:ijighly charged and ambiguous environment is likely to result from the circumstances~urrounding the activation ~
1. and the political consequences of the attributions themselves and the policies or actions that senior policy-makers may take as a result, it makes sense to focus upon building sound, pre-existing relationships among the relevant actors within the CJ community and its key representatives t<_> senior policy levels.
Only levels of trust built on longstanding relationships among the parties, as well as knowledge and awareness of the technical capabilities (and analytic processes engaged in) by the CJcommunity on the part of policy-makers, will help prevent dysfunctions in performance during an event -where all the parties have to feel assured that they will not be
"thrown under the bus" by the others.
## (U) "Safety In Numbers?" The Hurricane Katrina Problem
(U) How many people died in Katrina? Early on, estimates flowed in from a variety of sources; hospitals, morgues, law enforcement, emergency management bodies, news services, unofficial Katrina web sites, and more. Some groups tried to aggregate data obtained from others, counting the same cases again and again. There were complexities in determining what constituted the categories defining how the person was killed. Individuals who drowned in the flood, or were crushed by debris were certainly included within those categories, but it is unclear how those who may have been killed in a car crash during evacuation, or succumbed to their preexisting medical conditions in the high-stress environment were categorized and counted.
:~~..
_.,;~
(U) In an environment of information overload and time pressure, keeping on~~s;i~ye on the ball versus constantly playing information catch-up, is a challenge. Dependence:9]l..Jnm:"an memory for keeping track of information (for example, to pull threads or,:f~ll9w them~~)', especially given the acknowledged limitations of human memory anq c~gt;"liiY,M,iases;is destined to be inadequate, if not fail outright. From the teclm.ologic~:;I}~rspediW;<:tbe concept of an information clearinghouse is a step in the rigtli dix:~ctioii;i~!.JJ.an actual analytic
;::~:~:::::.:::ili~pro==~~\~::!~~-wuups Mil need a common framework, or a set of COmm9A:P.r!>..CeSs~~ir work, to help clarify that they are focused on the right things, and t~.afgaps~~~ beifi!ff]:~.~d. It will be extremely important to remain focused on the key .p:qin.ts:
;~::~
~Y
-
(U) What do we know?
. -::;;;:;:-.-.. /
.
-
(U) What do we need to know, but do=ii~g-
_
-
(U) Can we fill in the gaps? How?
-::::::::
... ;::~
.. :~-.
(U) This will require more than.P.~rson~_l~discipli~e. Under pressure, procedures, processes, tools, group norms, and commbti;~ituational awareness will be critical I
l Having a scorecard to fill out may h~lp_pr~vid some critical structure and focus in constructing an assessment. I
~
I Clarifying these goals/tasks will help to clarify ~~ttC flow/p(Qcesses/procedures, as well as help to define an appropriate analytic lll')~jJ1(Qm)itt~QI.J.. environment.
.
.. :;~:.:~.;~~=-=~~
==~
.. ::::~
~~:~;:
;~;;t:: ....
(U)""Tii~:~wr.ounding Psy~hological Context for JL._ __
_jiPolicy-Maker in WMD
AttribuiiOfi:!t
~ ..... ..
:::::~"'. ~~~
## (U) *Te"Orists Are "Adaptive Enemies"Jlfbo* Wilit~ To "M';;Fid/The Waters" To Complicate The Attribution Process By Jncreasing)Ifiiii'Fg!Liij!}' .
-::)~ ..
-:~ .....
In addition, inf~r;mation;,tliat evolves may have been purposefully Ld=is-t-ort....,--ed~b-y-t,--he-o-ppo-s-:iti-:. o..Jn, which mig~J;:irilslead 'investigators and conceal information about perpetrators and their actions._.t'T:~ftarists should be considered the ultimate iil adaptive enemies, who in the event o,fa WMD "ipcident will inflict damage (both physical and psychological), as well-~ g#tsuppofi:and admiration among their followers. In many scenarios we would ~nyis'iori'l
~
I terrorists likely will have a great interest in confusing 3.!1.<! _!::omptf~a.ting our response, and magnifying the psychological effects of the incident. :I
~-~~~---~-.~--....,--~---....,---~1 1nlSWOUlOOea~tCUlM1Y
important eop,'sJ9ef1ltion in the event that a state actor clandestinely supplied fissile material to a terror group;.:~~:f~rrorists may expressly attempt to leave an impression that more attacks Me impending in &der to generate fear in the public, as well as apply added pressure on policymakers.
## (U) Summary Of Recommendations
(U) Make I
I a valuable information hub for policy-makers, a one-stop shop for gathering the coordinated attribution information flowing from the three communities.
(U) Recognize that pressure for immediate responses will be directly proportional to the magnitude of the consequences surrounding an event.
(U) Recognize tolerance of risk by policy-makers (and their willingness to accept various measures or estimates ofcertainty) will be reduced as the magnitude of potential
.
SE~RN
cont::J(for either action or inaction) increases.
(U)
must calibrate communications with senior consumers based upon the way in whic
1 erent leaders structure their advisory systems. Understand that not all leaders are the same.
(U) Understand that people respond to estimates of risk on the basis of whether these are presented in the domain of gains or losses.
(U) If attribution information presented to senior policy-makers diverges from their preexisting, deeply held beliefs, more evidence will be required to gain their attention.
(U) Fori to get the right answers, they must learn how to ask the right questions.
(U) Use the correct set of expert advisers to interact with policy-makers, those who are
"bilingual" and understand the political/policy environment and problems faced by the White House, as well as the technical aspect of the attribution equation focused j I
## (U) The Right Stuff: Harvesting Expertise
(~
Having examined the range of capabilities that the US Government will bring to the issue of nuclear attribution, we conclude that IC, LE, and TNF capabilities, as currently configured, are likely to result in *eventual* success. By this we mean that we are confident that these efforts would eventually result in identification of those who mounted and sponsored any nuclear-related attack on the US or engaged in related activities. We are far less confident that as currently configured these agencies will be able to deliver meaningfUl, rapid success. In the event of a nuclear attack or imminent nuclear strike, the LE, IC and TNF communities each stands a cl).ance of uncovering meaningful and timely leads to. the perpetrators. Current resources might succeed in interdicting an attack or, aft~r an. attack, providing the President with a quick and accurate identification of those respcip~jole.
However, in such a serious event, it is not sufficient to say that these resourceS"7ii'jgfzt
~ -
d succee .
. .......
":::;;<=
'>~~::: .. :~ ~
"!~="
(~The key question is what can be done to dramatically increa!'l.~~ c~iik<l.~~.ofrapid success. Rapid success is important in any scenario, but vitiii in scer1iili.as where~:\Ve have an opportunity to interdict a nuclear device before detonation. W.<i:~P,nclud~;~~~u;_vestments in individual attribution capabilities are necessary but not suffj.~_ieiiti:9; gtake .li'ieal difference.
Rather, the greatest progress will come from forming a ~-~~fuless teaijt;.trom the IC, LE, aiJ.d TNF communities. We must tear down the re~ingJfilp~iments to"teamwork in this arena.
. -:::(::~:::::::~~J: ;:;;~\;;/
~>I
(U) In this example, a senior LE analyst is chosen to serve in a rotational assignment at DOE.
Following this tour, the analyst returns toLE-related duty. After a set period of time, the
. SECRE~ORN
analyst serv.es another rotational tour in an applicable IC organization, returning to his or her
~ffice with experience in both DOE and IC communities and better able to support L_jsues.
(U) C.~cting and Preserving Perishable lnformtltion
(U//FO 0) The information age has enabled a high percentage of citizens witJl in~tantaneous communi tions coupled with sound and photographic recording capabilities.;:O.able News Network (CNN) has exploited this new capability with their "!-Reports" in wlii(;li;:lilPyon~
who sees something of interest can submit it for CNN's broadcast consideration. Sfdri.e~that would otherwise be a mere verbal or written description are now docum~~~f~4UP. picblr~s and sound by the omnipresent public.
._.;.
}:;{
:::=::::
~-;-
. . ........
... ~
(U) In the case of a nuclear event, it is likely that individual Pri:~~~e citii~~~::~ili' have ima es stored on cell phones or digital cameras that could help the.attfiouti~n question.
should make arrangements in advance of any actual emerg~:~cy th~FW.~mld give e pu tc a way to send information to government serve~,f.o.r.an~y.j:~ .. During.iio emergency, the public could be advised via public announc.effi~riftPi::globiii::text message, of the way to submit photos and other information theY.;~ave tha:i:~ould sh~~~light on the event.
Technology now allows users effortle~IY:tb':;inct~de the date;. time, and GPS location in pictures. Experience with the Space Shuttl~:t"9l1,1lllbia crash debris recovery effort or the Hurricane Katrina response indicates a hugely.sXiPP.l!thetic public anxious to contribute in response to a national emergency/disaster:.::Ysing:ifils capability requires some advance planning, however.
--...-. :;::::==
(U/~UO) A second sourcc{~ij;irish!ible data is security cameras, traffic cameras, and clos~ddrcuit televisio~ {~X~T.V). If ~:;nuclear/radiological attribution event were to occur within a metropolitaQ a;iea of:i:be:C.oii'iinental United States (CONUS), a collection and review ofC9,TY da~~~ould beiieiii:.an investigation to determine the who, what, where, and when aspcrs::q:the cas:~:-~. Throughout most major metropolitan areas there are a multitude of publi~.~~d pri\ii((e CCTV:~sfems. Most CCTV systems are designed to regularly delete old footig~;~m !}i~lr memory systems. The value of CCTV information became clear following the bomoipg;pfthe London subway system on July 7, 2005. The subsequent investigation revealed sU'J;.~iA~ce footage of the suicide bombers parking their vehicle at the Edgeware Road subway~t~tion and entering the station. From the surveillance footage, the bombers were identified; greatly aiding the investigation and leading to discovery of the planning process for the bombing.
(U//FduO) A system should be implemented that will preserve all relevant CCTV footage at the very ~tset of a nuclear/radiological attribution event to aid the investigation. The CONOPS should incorporate a process whereby I
Foordinates with the Department of Justice to ensure that a formal request for the preservation of records and other evidence pursuant to 18 USC. 2703(f) pending further legal process is issued and/or the establishment of an agreement network through Interagency Governmental Agreements
"(IGA).
## (U) Information Sharing Among Disparate Kingdoms
(U//FO~O) Successful attribution depends on pulling together data from a broad range of governmental and non-governmental sources and applying the skills of analysts working across a variety of disCiplines and locations. A nuclear emergency would require that this work occur at unprecedented speed. More than any event we can envision, a nuclear emergency inside the United States would require that knowledge flow rapidly between the relevant responders. Current limits to information sharing exist for good reasons, including the need to protect sources, the need to avoid tainting legal prosecution, and the need to protect rights to privacy. These. reasons will remain important in a nuclear emergency, but c_annot be allowed to impede the higher priority of protecting thousands or mi:!f.~~~ of human
::&'0) By its nature, the normal method for cross-agency and diS!aJ!,~;~~~:is largel;~erial and often hierarchical process. This normal way of doi~!{b~~~ss is.'too slow and exclusive to bring the full capacity of the distributed law. enforct}fu,~nt, int~U~~ence, and technical communities to bear on the problem in the quick~St mtmn~;:~;~~;~ nucl~ar emergency, we will need these three communities (IC, LE, ang~":f,NF) toti~~cmt from each others' knowledge as fast as technology allows. We must prpar~~:i:f. tools~d approaches now, that when activated for a nuclear emergency, allow;.iei~vant j;ij~e~ to share knowledge at the speed of technology, not the speed ofbqr.~~q!;rac}i::~;::.
.:
(U) Capturing Low *Confidence Data and* Assessm~nts
(VI IF~) It is important that we protect against the loss of "low confidence" data and assessments that are easily overlooked in a crisis. Individual organizations may tend to dismiss low confidence data as "low value" data and not pass it up the chain or share it with the rest of the community. Taken as a whole, however, the collective picture,drawn by integrating all of the data- including low confidence data- may open up new;~~i.~nues of investigation or lend support to alternate hypothesis. Understandable organizati9P,~l aversion to risk should not be allowed to be an impediment to integrating this dat~.~.t:eal tlffi:~;!J!~o a single environment where it can be aggregated with other information lP:iifJiC:.~~ibly prbvide
. '
~ ~-" ~.
additional value. However, once the augmentation group i~fonned.!!AcJ connec~~~i;-it is essential that low confidence data, with proper caveats, be :;;c>ntinua:ily:]}iitegra~d;~into the assessment process.
:::::..
:::;;:::/
... " =::::;:..
.~~:'
... ...
'""'~ ...... ~.
~ :~~.
.. ::~:=~-" .
(U) Continuity of Connectivity
.. ..
.;:;::::~.
/.
(U) The connective process should be ~i~b~!~~~~~i~:~~-~.Jow level with regular alerts and exercises to test connectivity, trans~!i.<!q. . to o.ry.itional st(tiiS, and to encourage the community to be interactive. In effec~ this::~B.~.l.d;be continUally building and improving the system in anticipation of an event. Through p~~ipation in test alerts and transitioning to operational status exercises, c=Jnembers wiil=:~M:amiliar with the computer systems, applications and resources before a crisis ~~ 'ijiis type of approach has proven very successful in the NEST Program (Nucl~#;Emergfncy Search Program) .
.
-~~=:~:::::-
-~~.
(U) Once the tools and proto6ols,for the larger expert group have been established, the concept could be expan4 tQ:a.mucJ:d~rger group of potential collaborators via parallel communication and ~\iiit{l~Y.rn~gfbut with the same focus of real-time participation and information ~b;uing:(~~'~ppropnaie to the community of interest). In this fashion, the huge resourc~;efUQ.'~~~slfied~WOfl!lation and brains might be engaged in a "hive" fashion that coul<tJie mini.ti"g::infonnanQii"and investigating hypotheses in parallel with the group selected I
I ~q:\vork in thl secure environment. This unprecedented engagement of
. informati~~i:~sources would be transformational and should be a consideration aslr------, augmentati<f~i~omponent becomes a concurrent collective and lessons are learned.
-=!;~~~= ..
(U) Leveragi~l;~e Catalyst Project
(U/If~UJ~1\11 11------------------------------------------~~
| (U) | IT | Interfaces That Enhance Interaction |
|----------------------------------------------|-------|----------------------------------------|
| . | ,:; | . .{ |
| :::::;.~-- | | |
| (U) Information Technology (IT) is integral! | | |
| .~.: | ~_as | il~~~iJjtates |
communication and ultimately enables collaboration to produc~;~ffective'and'timely analysis.
In order to leverage the benefits of the IT available j
. p.i;~p be .iiiiportant to plan and coordinate these tools before the start of an exercis~,~f an actmil~~s"is. There are two primary areas of concern in the use .of IT as ~-~M~ler-6~~!?-~municationj I
| 1. | Data sharing via common databases, | anQ.:::'" |
|------------|--------------------------------------------|---------------|
| -~<::::~: | -;:::~:.: | |
| ... | ,. | |
| 2. | Information sharing via social networ.\#g. | :::;: |
| :?: | | |
| .-~ | | |
| - | - | .. |
| :~:::~::~ | | |
| _ | .. | ! |
| .~:-' | | |
| .- | | |
(U) Each of these areas shares some common:~~~llenges with regard to successful implementation:
-;~~;~~;;:/
-
Particular clearances are needed for.:~~~~~sing agencies' data.
-
Computer systems that c~:~rate':at a common level must be properly cleared.
-
All members 4
, i IJ),U~t be (amiliar with the computer systems and applications
before the start o.~ ~:~~s~f:;:;::::::::}
(U) *Data Sharing* V,iil.~Common Databases
(U) A IJ.w:tilie~~~~:o'~g~i~tiQ!lS have proprietary interests in their particular databases. In
ligh~.~~:!}lis, it Wadvisabi~::fo~ the National Counterterrorism Center (NCTC) to maintain a separi:ifu:1U.s!~~g'of other existing databases that can be accessed through assigned points of contact (li#~-?.!1 officers).
(U) lnform~i~~;S~aring Through Social Networking Tools
(U) Tools are'~vailable now that, with some work, could helJ1r--------'ir,----,l
'-:--:-:--:-:---:--=----=---=-_.,., This widerc:=Jworking group could consist of sevefal hundied individuals from the intelligence, scientific, and law enforcement communities, all needing to share the same data in an integrated workspace but far from a common workplace. One possible near term approach to this issue could involve the use of existing classified network tools such as lntellipedia or A-Space. Special access r--liostances of these useful tools could provide properly cleared analysts with a place tOoiSc'uss ongoing nuclear attribution activity, share assessments, build consensus and develop alternate assessments. Use of the existing infrastructure could provide a rapid. relatively inexpensive way to share information and build a more integrated analytical community. Once piloted, the infrastructure and tools might be extended over non-traditional elements such as state and local law enforcement and
the broader technical community.
(U) There are many advantages to the use of social networking tools- Facebook and Intellipedia, for example - as opposed to the typical way of delivering data in the form of formal reports or serial distribution. One advantage is that information is organized topically as opposed to organizationally. The use of these tools allows for a collection of data in the form of what is known as a whole, rather than what an individual organization thinks based on its limited set of data.
(U) These tools also have an inherent ability to record the development of ideas on an issue through the archiving of previous assessments and the discussions that led to those assessments. The effect is to allow rich local data mining and a high degree of concurrency in sharing information and interaction, which can also serve to increase the lik~lili'ood of interactions across comniunities, a somewhat rare occurrence today. More int~ra'9!i.on wjll help to create pathways that will improve our ability to harness the colle~P.~~ braiiip9~~j. of the institutions involved.
}'~,;::rz.;
.-.
...... :
::~:~ ........
(U) The Intelligence Community (I C) has established a nufu~er of ri~~:~r:ogranl~imed at bringing "Web 2.0" technology to the Intelink user. "Web 2.~::i:~~Jhe sec~M-generation of Internet development and web design. It is characterized ~JaCiH~fulg coi:iimunication, information sharing, interoperability, user-centered des~gri:and colia'69pttion on the Internet It has led to the development and evolution of}v,eb:-basid:communitids, hosted services, and web applications. Examples of Web 2.0 enabl~;~~b.d~c~~:irlt:lude social-networking sites, video-sharing sites, wikis, blogs, mashup_s~~:and fo~onomi~:~;~:,;
(U) One of the key concepts of Web 2-~~.--~~~~t~~ to Intern~~ experts John Batelle and Tim O'Reilly, is that customers are building your bus].i!.~~~.for you by generating content- in the form of ideas, text, videos or pictures- thl!-t~ b~:Kamessed to create value. Users can own the data on a Web 2.0 sites and exercis~.:~6iitrol o'Ver that data. Web 2.0 sites encourage users to participate and add value to.,tb,e;:appll~ation as they use it. This differs from earlier approaches, where content "'~ ~tatic, vsers were merely passive viewers of fmished products, and only the *e q~h co~~ change the information.
!=~:~~
:::~;.::~=:~::='
(U) "Socia_l -~dia"j~~ be thought"ofas a fusion oftechnology and sociology, where modem technolqgi~~:m:,\lsecf'to;:pring people to connect and form relationships for personal, politi.~~~ busin~~s, and j)fQ'fessional uses. Web 2.0 technologies allow for a range of Social M~di~;~~~h ~;:ihat found')n the Intelink applications A-Space, Jabber, Intellipedia, classified blogs, ali":~qtJier applications .
.
-~~:::.: .. .
.. .. -.... .
(U)/
-
-
-
(U)
. -
(U) InteUipedia -This application is, in effect, a classified Wikipedia. Users are not
anonymous and are able to create or modify web pages covering a wide range of topics.
-
(U) A-Space - Analyst Space is a social media application where users can sign up to
discussion groups and share information (including finished intelligence, all-source
intelligence, open source information, and other types of media) with other users
throughout the Intelligence Community.
-
(U) Jabber, Intelink Instant Messager- These tools allow users to send short messages
to individuals or user groups in a secure environment.
## (U) Recommendations
~..
.. ...
-
(U) Establish a concurrent work environment for all those who support'---'"""';-::-----'
I
~hroughout the country and in law enforcement, intelligeli't~.?<and -
technical communities (as in A-Space) to assure collective awarenes~;,jpteracti'sQ.;;!l#d
integration of diverse data, including especially low confidence da~rz~;::;~~.
.,
| '~ |
|------------|
| ... |
| (!~;:::;~" |
-
-
(U) Provide networking equipment and ca.n~l;lil;tieS:~t~~:ickly ~~:~~~arate communities
into a concurrent communication envir!)nmeiiltthrougil;c~P.~ilities such as a STE phone
and media-less terminals)
._3!:;..
:~;::
;;:
-
(U) Use social networking tools to ~n~~~1~~;~;;~ective ap~roach of analyzing data. Test
and further refine the network!
ftlii-ough regular exercises and real events
(at lower concern levels).
~:~:-
~ ..
- M Test the possibility ofeijijitid~~t~e collective beyond cleared and individually
authorized people to the;~iB!'ger coriununity in the unclassified environment to evaluate the
benefit of engaging;tjps :~Hge resqiP"ce .
.
~. ::J';
~~:~:~:::::~:!~~~.~
-
(U) Ct:t~~t~:-WliY~Q:\,~tandards for analysis coupled with a common protocol for reporting
pUJ"P.O~~~=,!~T!i~s woiii9~~~rye well in minimizing tradecraft differences among agencies .
... ::::..
:::~
~:::r
-
(U):;r-----ri~formatiori'~haring should take place within a common computer system
inte~teared to the highest practical levels. However, given proprietary data among organiziti~~.- the NCTC should have a key to databases with assigned points of contact,
through w~ith access to individual databases may be enabled. The chosen fl
computer Infrastructure should be configured as soon as possible, and all c:::Jii'embers
should be rapidly yet thoroughly familiarized, trained, and exercised on these systems,
applications, and protocols.
## (U) The Mixed Reality Perspective
(U) This article was provided by members of the Mixed Reality SHARP group, which was meeting in the same facility as the Nuclear Attribution group. This article discusses the establishment of an infrastructure for supporting anc:=Jcommunity of interest, which will facilitate rapid analysis of an impending or recent event.
## Analysts And Law Enforcement
(U) **Intelligent** Ageqts;i;~;~
.
.
(U) The ~~btl~lo~l~ncept for streamlining access to pertinent data, and facilitating commUP.icafici~:;aeros~itiJ~c=Jcommunity, leading up to, during, and after an event involY.~~ three iQiijor compdnents:
-
(U);~~~.at~ess,
-
-
(U) ~;!B~lligent agent-based system and
-
(U) sectii~;tm(jbile hardware .
- 11'1:&'1!'
.. :~'!
(U) Overcoming Classification Barriers
(U) Members of the CJ community should be provided with all of the data they need .
when they need it. Each piece of data should be tagged with the level of access required to view it. Tagging each piece of data will provide granular control over access to datasets.
Rather than applying credentials to an entire dataset, which may exclude a user from all of the data contain~d within, tagging each piece of data ensures that users with the appropriate credentials will have access to the entire subset of the data for which they are cleared, even as their clearances change.
(U) Enhancing Data Mining
(U) Software robots known as intelligent agents should be utilized to access and collect data on a user's behalf. By passively learning from users' behavior over time, the intelligent agents will anticipate the types of data to collect for each user. The robots will be able to access and store data beyond the users' level of clearance and will act as a proxy to this data so that all necessary data will already be in place for when a user may be granted more clearance to the data. Agents also will have the ability to proactively notify users that there is data pertinent to their interests to which they may not have access, and will tell users to whom they must speak to gain access. This approach will ensure no user has access to offlimits data prior to being granted more access while at the same time ensuring that the user will have access to pertinent data as soon as it is deemed necessary and properly granted.
(U) Secure hardware dev_ices should be put in the hands of the r--lcomm11D:~~~;~~~o that i~
members can access pertment data, and each other, whenever ~erever reqo~~
.. U~mg this technology ensures that local law enforcement and on-the-ground ~~~n.nel "'"i4~P..~ve upto-the-minute, secure access to intelligence, and may in turn provide int~iiigt:fi~ and
situational awareness to community members who may not,J~e on-tq~~g{oun=;ij;n-<iware exists today that uses mobile technology similar to the BlackbelJY In cenj:qnction .with cryptographic devices to ensure secure communications. Lap!.O"Ji~:a.lso ~~;~yahable that provide the same capability but are slightly less mobile. II].:;qlese''Si?.!"!:\~iol;ls';'data security is maintained by serving data from a remote server withmJtstoring any:9ifia locally .
.. ~. .. ",.'
I .
.
.. ;
~ :~:~; -:.. .
~~ ~~:~: =~~~..
.
(U) While the above discussion centers on~te'Ciinolegy, esci:b\i~Ji.ing an effectivec=J
network relies on a solid community o9~~sted .ilSers. Thi~:~ommunity culture challenge
requires an appropriate medium for key melj:l~~(Sio identify each other and discover
individual expertise and interests. Regular c6fiiil:tunication between members will contribute
to the success of the system.
'::::~:::::
(U) Of note, the approach described ab.g~;~um;~-~t data in the pertinent datasets be pre-
tagged with access permissi~}\l::r.I@ts, iequirements, and metadata concerning the data's
content. While this is a tracthbl~:Yet O:ontrivial task, further discussion is beyond the scope of
this document. Similarly., t~~C:=::ft-~iwork approach will require the sponsorship of senior
leadership to addres~_poltcy iiifd:atS'iliiizational challenges, which is also beyond the scope of
this appem~i?t:..
/~:t.
. '"''"
.. ;://'=-=:;;~j} . ::::~=~~:;:-
(U):by:iilWJic ~alysis Process foriL ___ __,
(U) Verj:~lk4 to nearly intractable intelligence problems usually go through an initial "stab"
by a highlyr~~.d special analytical team. Outputs from the team's efforts are used to
update collectiQii' requirements and sometimes result in the formation of a new entity to work
the problem fuil time. After a long period of little progress or no real intelligence production,
the problem may fall below day-to-day focus. If and when activities related to the concern
about the issue remind consumers of the threat, then another focused look may be initiated.
(U) For certain classes of issues, failure represents catastrophic consequences. Issues of this
nature require a level of focus that demands deep daily digs, even if the result is negative over
long periods of time. Here, dedicating some resources to working the problem on a
continuing basis is worth the effort. A dynamic approach to analysis and collection for hard
to nearly intractable issues is illustrated in the attached chart. 30
## I
(U) The dynamic analysis concept is straightforward. The first step involves creating a notional series of phases that any entity would use to start and achieve a particular objective.
The actual steps would be based on whatever is known about the entity's ilk and the real world activities that would have to take place to achieve its objective. Under each phase or step, participants in the process brainstorm the manifestations of activities tha! m~t occur for an entity to proceed with its objective. In concert with this brainstorming, patii~_ipants identify what might be collectible from any particular manifestation. For exailipi~;:.there_
would be a number of areas where information and phenomena would be ge.nerateii:":i11.th"e course of arranging for a safehouse for a covert meeting of the "group-wlio!l\~t_f)s-potiiioes."
Someone would have to make the arrangements for a roo~ P.r hous~,:~tli-ereby 'Iiljing a paper trail. Some attendees might rent cars, another paper trail. ears on th~i~:way to .tHe meeting would cause Doppler shifts in local radio frequency signals. Aii:upswing;iQ.~ti-potato rhetoric might appear in internet media. Collectible info~Jibti-~9:t:ild be:fu:.rayed against the capabilities of existing collection resources. If a collectioii"iesource:;~~"rthe access and ability to collect the information, an appropri~~t:-r~ui~~DJ~Q.t would b~ initiated.
Alternatively, requirements already on the b.o"cik's':wpuid be:lij,dated if necessary. If collection were feasible but no capability yet exist~~:~ ass~iated gaj;:::~iosing research objective might be initiated.
/
::::~~:::; ...... :
.'
-:::-.
(U) When run on a continuing basis, the dynami~::~~lysis process would constantly iterate the issue's evolutionary steps and possibl~.p;tanife"silltions. These iterations would be based on any enrichment in understanding th~-:~'ofthe'target entity and specific intelligence information. A simple analogy..:is;~he cb'iistant process of writing and rewriting scripts for a movie as a movie director sltoo~:ilild e.valuates scenes or as the scriptwriter does more research.
..:::. \:(
.. ):
(U/IF'n.uO? _W~r~ ?-~;J~~lear~~~t~~n is more than just picking up the ?ieee~ after an event takes p~:::~tgtlance:ui-:workmg the tssue before any harm takes place IS a v1tal part of the job. ~!l.i~rdictl~~;as eariy=~-possible should be the primary objective. Substantial resources shouia::ti~.~ed.idlited to fre<juently revisiting what indications might emerge in LE, TNF, and intelligeii~::st'ata from motivation to the aftermath of a terrorism-based nuclear event. A
dynamic aii~JY.~!~ P.rocess should be among the responsibilities of the dedicated working group propos,1for the c=:FONOPS.
## (U) Leveraging Local Assets
(U) l"'reasing Efftciency
(U//FOUQ)I
| (U) |
|------------|
| ,.{~;~ |
| ::::l:: |
(U) Local law enforcement may be the primary means of _q(ili~tirlg:;q. intelligence in a domestic nuclear event. Crucial initial information likely.:will be deheo from local law enforcement. The Intelligence Community mU.sHmsur{itiii_tJocal investigators are seen as valued contributors. The best means of a~qornplfs~{!lg this:i~:t,ht:ough proactive teambuilding with common training as a basjs";::..
.::;
-;:::.
... ..
~ ....... ": .:.
~-~
.
(U) The Bush Administration's "Natio~al ~~~~~'for Information Sharing" (October 2007)
directs the continued support of fusion centers ~i:~ij~~state, local, and tribal levels. The following considerations are suggested al!-:W.eans of enhancing this crucial area of the criminal justice system.
,::;;:'
1. (U) Aggressively foster. c;:xj)S.:Q8ioitof the network of state and local fusion centers.
Combine the efforts ofth~;F~I:Joint;Jerrorism Task Forces (JTTF) with the efforts of the
local fusion centers l9;.~~nli~~e the. ~Q:fonnation sharing process in venues where both exist.
Each agency 1n eithefisr txifli::a-;H~iF and local fusion center, within the constraints of
mandatec;l.it}{Qrm~ti~ij classifl~~ti~n restrictions, should be aware of the collection needs of their paffii~i;~gencie8;:3;:.;.. .
.... :::
\~::~
.. ::;:Y'
2-:(U}~~Q~jd~ technica{tools and training to state and local law enforcements to enhance their a&il}ti~s to recognize, collect, collaborate, analyze, and share information. The training sho~l~. jnclude a standard training protocol for LE to become familiar with Federal guidelines oifiiiformation sharing, investigative techniques, and source development focusing onintelligence and terrorism data recognition and collection. This is effective not only in combating terrorism but likely will result as well in enhanced crime reduction.
Historically, local detectives have done an outstanding job in developing street sources and criminal intelligence, so additional training to hone these investigators' abilities to recognize and pursue terrorism related data would result in potentially vital contributions.
3. (U) Develop and implement a nationwide intelligence liaison program linking federal/state/locaVtriballaw enforcement information sharing activities. The model of the Terrorism Liaison Officer (TLO) - developed at the Arizona fusion center- should be emulated nationally to support intelligence operations. This TLO program has proven tobe an effective tool for disseminating informatio~ to all levels oflaw enforcement, and to entities within the Intelligence Community.
## (U) Arizona's Tlo Program
(U) The Arizona fusion center created the Terrorism Liaison Officer (TLO) position to address the center's need to reach out to new enforcement entities throughout the state. The TLO has two intelligence-related duties. First, the TLO serves as a conduit between the fusion center and the TLO's home agency. In this role, the TLO identities and disseminates pertinent information to his home agency including street level peers. Privy to the fusion center's collection needs, the TLO monitors his jurisdiction for this information and advises the fusion center of the relevant information. Second, the TLO serves as the "eyes and ears" of the fusion center when responding to an incident. The presence of the TLO on-scene provides observations, reporting and validation of information from a vetted source. This is invaluable for reporting on real world incidents and also for preventing unnecessary escalation for non~vents.
(b) (1) (b) ( 3) (b) (5)
(U) CAN OTHERS UNDERSTAND YOUR DATA?
## ~:I
k---------------------------------------~
(U) The following provides a framework of four general scenarios of nuclear/radiological events: pre-detonation, post-detonation, pre-dispersal, post-dispersal.
I. Scenario of Pre-detonation: Investigating and attributing a threat of nuclear detonation or interdiction of fissile material.
2. Scenario of Post-detonation: Investigation and attribution in the aftermath of a nuclear detonation.
3. Scenario of Pre-dispersal: Investigating and attributing a threat ofradiological dispersal or interdiction of radiological material.
~--
..
4. Scenario of Post-dispersal: Investigation and attribution during or in the~'~ftermath of radiological dispersal.
;::-::~:~:~:-.
.
| ~ | ~. |
|------|-------|
| .., | . |
| ~~ | |
(U) The following best practices are recommended:
}::~:::::::,.
-:~;::-:
-
Maximizing information sharing and collaboration in. a connect~d env'itdiiment where
"need-to-know" and security issues are remedied to=ffi.e ~xt~P~:s,sjble .. :~:=
-
Erring on the side of retaining and preserving informati~Q~,.includi~g~low confidence
information.
.
~~;:::.~ ... .:;::;~::-.. _ ... ~.-
-
Applying analytic methods such as the Dynami(f.A.jialysis Prq~ess and Alternate
Competing Hypotheses.
_.:::-~;::::~j._~_;; i'''~~:(;ii: .. _~.!:::
.:-
.-:::::~~
(U) Analydc Methods (U) Dvnamjc Analysis Process - a "cradle-to-grave" methodology of activity awareness (see Appendix C) (U) Analysis of Competing Hvootheses - an analytic tool for comparing alternative explanations, where all conceivable hypotheses are analyzed in a matrix against corresponding evidence. Each "hypothesisevidence" pair is weighed to develop a probability or confidence level for each overall hypothesis.
-:-~ ....
... :~~~
:;::~--
.:.~-:.
## (U) Data ~~4~1Ing::~~;'Vis~~~:~~~N
(U/IFCWd}'X:~~y-to ~~~~tpg;any crime or mystery is the employment of effective methods to gath.~r;;jpmlyzei::llnd relat~;mformation about the case. In the process of collecting and fusing intelilgt}:qf.t?~.-~G ability to. visualize numeric and non-numeric information spatially and temporafl~i~-paramount. Laying out the information in graphical and/or tabular form, with associated 26M<Jen:ce levels, and showing links to locations, movements, materials, actors, nation states, ~jf:; can provide a rapid relational context. Maps may be utilized to show not only location~fbut concentrations, correlations, directions, distances, domains, flows, routes, vicinities, and countless other geospatial relationships. By the same token, event time lines can be graphed to visualize time sequences and intervals. Additionally, the temporal information can be mapped to graphically show the effects and nuances of evolving events.
(U/~TTO) Insofar as efficiencies are gained and time lines met, the ability to graph and vi~~~ay include the full array of classic information visualization methods, such as pie charts, bar charts, and scatter plots as well as relatior=:J: graphs, ephemeral "flythrmighs," and virtual renderings of scientific data.
should employ data visualization techniques that enable robust analysis an o er tunely illustrative presentation graphics that accurately portray current analytical findings.
(U) Using Metadata to Foster Information Sharing Among Separate Communities
(U) In order to facilitate ready sharing of information among well-established communities, metadata could be used to ensure that community-determined relevant inform;t.tiol). is available to all cognizant parties involved in the attribution process. By lever~gwg metadata successes implemented by other communities, the use of metadata would facilita(~:ld .
accelerate effective data and information sharing among the nuclear a~.~~t~~n co~~ities as designed and developed to meet each community's needs.
} ,:::::<
-
~-~ :-
~~~-:-..... -~
..... -
... .:~
.......
. '
(U) What Is Metadata?
(U) Metadata typically refers to distilled infonnation that describes the semantic content of a piece of information [I]. An example of metadata in nuclear forensics is the geometrical parameters of a cluster of isotopic measurements (e.g. major and minor radii, standard deviation ellipses, etc). A cluster may comprise thousands of raw measurements, but the geometrical parameters that enable the cluster to be defined are the metadata. In support of data interpretations, one would match a questioned sample with the signature families that are denoted by the cluster "cloud," not with the actual raw measurements, i.e. match with the signature family, not the data points.
::~:=:...
.
(U) Metadata is formulated within each cornmuiiiij:by their experts, and the originating community determines what information.~iifi be distilled into metadata to be shared with others. The strategy is that cornm~tic{will drive the metadata development process so that they are comfortable with tn~:iti~tiiaata.constructs that will be available and shared. Metadata also is used in directing datal.apd "inforination searches into a manual search mode using simple contact infonn~.~pn ~~4~-- .f.pf::Cxample, a "library card" with contact information would be inserted iqt.Q;a metaClata:<lliiabase so that an analyst would know whom to.contact for more i!J.:fpffil!ltioii-:fij{l! may be relevant to his or her task, when the additional information catino~.;blsliar~;.automilft~11Y via a database:
(Ufj~~~!l~.iil~~ is used~:~~ facilitate and accelerate knowledge discovery among disparate collectiotis~;(if.pata and information, as it provides standardized data constructs for latent pattern analy~!~:;;:For example, the MASINT
31 community, via the MASINT Standards Working Group:"ofthe National MASINT Management Office (NMMO), has deployed common metadata templates for all MASINT executive summaries. The use of common metadata templates enables fonnulating metadata from the MASINT executive summaries to aid in finding commonalities and hidden themes in the MASINT summaries.
(U) Similar to the MASINT community, many other disciplines are using metadata to enable effective information sharing within their community and also across separate communities.
The following general recommendations have been identified to facilitate successful implementation of metadata to enable effective sharing among communities.
- (U) Explicitly designate ownership of a classification: whoever is the original source of the data is always correct with respect to the application of the metadata to its
## Se~
description. 32
-
(U) Ensure agreement between researchers and domain specialists: although researchers
may be more consistent, their interpretations of the metadata may differ from specialists
who have direct experience of the specific domain. 33
-
(U) Ensure consistent quality of meta-tagged data items: it is important to refine data and
information that does not p~ovide sufficient contextual information to enable metadata to
be formulated in a consistent manner. 32
-
(U) Ensure consistent training and skills of the metadata developers: formulating
metadata from data items has been shown to be strongly influenced by preyious
experience with coding.
(U) RECOMMENDATION
.:::::
i=:{::... ::~)::-'
(U) By leveraging metadata successes implemented by other c~mmuniti~sl:~e use of metadata would facilitate and accelerate effective data an4,:ipforrDB.:tiQ~ shanng among the nuclear attribution communities as designed and develQ~ifto meet'~~cli" community's needs.
## (U) How To Best Use Tnf Information
(U) Managing Expectations: What Can We Expect From TNF and When Can We Expect It?
(U) A post-detonation IND event can be accepted as having the most dire consequences for the public, and, consequently, as posing the greatest pressures for timely and accurate attribution. Therefore, this section presents suggestions for best practices and other considerations for managing expectations for just such an event. These suggestions can be adapted to improve expectation management for the other nuclear attribution scenarios, such as a pre-detonation event, the interdiction of an IND or-an RDD prior to detonation, the detonation of an RDD, or the interdiction of nuclear materials where no device is evident.
(b) (1) (b) (3)
(U) Recommendations
-
ill) I
## (U) Limiting Bias In Technical Data Interpretations Secre~
(b) (1)
(b) (3)
(b) (3)
(b) (5)
## (F~O) Put'1;Jng It All Tqgether: Proposed Function And Structure!
b)
(3
(b)
(b) (3)
(b) (1)
(b) (3)
## (U) Building The Nuclear Attribution Community 2030
(U) In examining the nuclear attribution problem, SHARP highlighted the fact that doing attribution well requires a broad and robust community of experts working together seamlessly, but in reality nuclear expertise is often sequestered behind physical, virtual, and policy walls.
(U) To improve information sharing and create a true community of analysts, a variety 'of traditional and modern approaches can be used. None of these approaches are cost free, and all require overt action and endorsement by senior management. However, by.stp:ting with a variety of approaches and scaling up those that work, effective methods can b~t;(ound. By building a community, nuclear analysts and scientists will benefit from buildiliif~~M
strengthening collaboration across the wider intelligence, nuclear forensip;.:l!J!d la\\l.~:::;~,:;i~
enforcement communities, with an attendant improvement in their analyiiS:::;:::::..
,.s~
.;;~-:~:;~
:.:~::~~:.:::
(U) Emerging computer technology along with traditional ~~nt9ririg''ciiii:-qe merg~d to produce a robust, distributed community of nuclear experts in~HQked tflt~ygh social networking. The goal is to achieve four major objectives:,.;.; .. ::::::;..
.:"
-~~ ::-
~::=:~::~
-
Enable Knowledge and Information Shruj~g;:;;..
)!:~;~:~..
::
.~ .... -~- ... . .
. . ~~'"
-
Bridge. the Nuclear Expertise Generatj~n'Gap:;:;:::
:~~==~=:~..
-
Sustain a Common Lexicon
. :::::...
.:::
;::,.
- -!"
.
,
-
Implement Virtual Exercises
.
''=::::!: . ,/
.. : ~ ..
- "'! !"~ -~
... : .. ::.
.
- .. ;.~!" .... "
.. -z~t~~ti-
(U) Trade Shows: Virtual and Real
.. :~;:..
.
(U) For the nuclear community ~Q __ over~ffie geogrnphic and organizational sharing impediments, it is necessary ~:iJeyelop"real world and virtual professional I social information sharing environm~t5: Nuclear experts are not immune to the human tendency to associate only with indiv.i.Prlils.or co.-:Workers with similar interests. The blending of social networking technol~gy;~nd ~~Jili~~wl or annual nuclear analytical workshops will create a communifX8f~s!;~~:!~nd itseiitoward a collaborative environment.
.~;4.:-. ;~~;~:=!~
::~;~~:-: ..
(U) A;ieai worl<l'example\~f such a nuclear analyst network is the Project on Nuclear Issues
(P0Ni:t~"<>~tcx~~&y the Center for Strategic & International Studies (CSIS).34 The goal of PONI is'fiij(;~~!lild and sustain a networked community of young nuclear experts from military, miii~~-~.}aboratories, industry, academia, and policy communities." PONI hosts four major corifeiences where young experts across the community present their ideas on issues ranging.from US nuclear weapons stockpile issues, foreign nuclear programs, to technical nuclear forensics and attribution. PONI also hosts smaller events with guest speakers and maintains an online blog.
(U) Due to the public nature ofPONI, there is limited involvement with the Intelligence Community. All discussions at the conferences are held at the unclassified level. PONI, however, serves as an inspiration and a possible forum to model after for the USG nuclear community. The proposed USG nuclear network will provide intelligence analysts with similar opportunities at the appropriate classification levels.
## (U) Recommendations
-
(U} Develop a nuclear analyst social network, where nuclear analysts from the
intelligence, law enforcement, and nuclear forensic communities could participate.
Rather than a workjng group, this is a support network designed to bridge the
expertise and age gap within the nuclear community. This could lead to the
development of local "chapters" throughout the US or subculture to the analytical
community.
-
(U) Create a "virtual community" of nuclear experts by leveraging emergent social
networking technology. Regularly scheduled informal chat sessions can be held on
the appropriate computer systems. The only limitation in this scenario would be
access to appropriate classified computer networks and systems.
o
(U) Such a community could include the use of classified or unclassified
networks to conduct meetings using privately hosted sites sucl\,as ~econd Life.
o (U) Classified seminars, briefings, and exchanges could be orgpj:lj~ed using A-
Space or other social networking tools.
; ::::::..
-
o
(U) Virtual world platforms, such as Second Life, could b~,;q<;lapted'ilr~:P.Qrted
to SECRET networks or JWICS to conduct simulations ~t'ili~:~JiJ:Ssifiealevel.
.:::::
.l~t-.
-:~:~~::/
## (U) Bridging The Nuclear Expertise Generation Gap :~:~. -:::~---.'"
)~::~~==~-
__.::"'
.
.:.-.....
(U) There is a recognized and widening experience and,ag~~~ap ~~~en' senior and junior analysts in the federal government, a gap that.iS.-P.l!J.St e\:iitJ~P! and critical within the nuclear community.
.;~::.:~.~ ~~=~~~=~~=:
t,::;~:;:~ ... ;
.
~ :~:::
:~:::
~,:~~=~.
(U) The number of radiochemistry pr()giifrii:s:{t~dikfio chenitsts in United States National laboratories and universities has dramatically~id~Jined over the past several decades. The narrowing pipeline of qualified people into this-~ii;tj:~al field is a serious impediment to maintaining a robust and credible nuclear_fQrensich)rogram."35
-
-~., :i:'"
-
r
.=.r:.~
(U) Senior analysts who are ~~Pii;i~nc~1' in communicating technical information to senior policy-makers are retiring. Tbi;;:is a sk;illlearned through experience and lost to the next generation of analysts as.,seliior.analysls retire.
~. ~=:~~~~
.. ~:;==~~:;:::~~::
(U) Recom!IJ.~!':datiO'fi_S;.,
-
OJ)'llil#~~ a pf~~-where junior analysts shadow senior analysts at high-level
.. ;:~Hnitiativ~~l
:
Ito provide the next generation of analysts the experience
- '::;fu;~e-~~liiping and'~i-esenting assessments to senior customers.
-
(tJl:~.p a regular basis, provide information to junior analysts on current assessments,
techlii~V~telligence gaps facing the community, and lessons learned from recent
assessriieiiis .
..
-
(U) Initiate training programs to instruct junior analysts in the tradecraft of writing
ahd briefing senior customers.
o
(U) This could include training with Toastmasters, or other public speaking
groups.
o (U) Specialized training to teach junior analysts to write effectively on nuclear
issues.
-
(U) Initiate a training program, similar to an existing IC initiative, where junior
analysts work and interact with scientists and analysts at the National laboratories.
-
(U) Establish a formal nuclear mentoring program that pairs junior analysts with
nuclear experts both inside and outside the US Government that includes formal/
informal training programs and si~sits.
## Secre~Forn (U) Communicate Today; Still Communicating Tomorrow
(U) As discussed in the Communications Section of this report, a common lexicon is required to ensure key attribution conclusions are properly formulated and communicated. The entire nuclear community will benefit from a common lexicon to communicate conclusions and associated uncertainty. A nuclear community network will foster the sustainment and adaptation of the nuclear lexicon for future generations, ensuring continuity of operations as the attribution community grows and matures (see the article "(U) ORGANICALLY GROW
A LEXICON" for details on developing a nuClear attribution lexicon).
## ( U) Recommendation
-
(U) In addition to the recommended monthly secure web-based, yjrtQally-liti{{!:t:;miniexercises, encourage the virtual coordination of analytic produc~Hiii=fi!l,~ tecilliology,
such as Intellipedia, "table top" exercises using virtual worldS',~~ihcorpof.itfug mixed
reality features in crowd-sourcing interpretations, Jith tpe o~o~~~f;J~ucle~ community
to foster sustainment and adaptation of the nuclear attril>iltion lexi'cant
;::::= , ... ;:~:~!!~::::~. .::
## (U) When Reauty Costs Too Much, Go Vi~~L:... ./~~{., _;::
.-/~.~-;;~;=:~:~:~: .
':::;~::;;:.:
(U) Conducting nuclear terrorism exe~j-~.9an beZ~kpensive:.~fiTo mitigate the costs it may be possible to modify or adapt commercial so'~Ar,e'to allow nuclear analysts a new way of participating in exercises related to nuclear afii!ibution. Platforms such as Second Life or a , modified "first person shooter'' video game col'ii(t:~e..used for scripted exercises. Analysts at remote sites around the world could log in.:t:P a secfue network and interact with their peers in a variety of nuclear forensic scenarios ~p~ii' as:
-
A pre-detonation sear~li~f;nu~f~ar materials smuggled into the US.
-
Post-detonation gath,~ri!llof it\rormation leading to an attribution assessment.
-
Operations at a,;~~(.opc;:~~i,ns center" where inputs from the field and
requirement~;;f.tom seme~::<fecision-makers are dealt with in real time .
.. ~;:o:::i:;:;:~~
_.:::::~::.~ ..
(U) Adyantag~~:t~f virtilil;l~xercises include:
!!::::~~.o mt=~equirel~~embers couid participate from home or their home offices,
' :-a:~~Q.dlng on the ievel of classification.
-
ln~~p~sive_ - limited software development costs, no TOY costs, limited IT support.
-
Allows=.forreal time interaction.
-
A vail~E:l~ any time - could have multiple exercises every month if needed.
-
would build team cohesion - learn how people work in a simulated crisis.
-
Troubleshooting- issues regarding protocols or procedures can be identified and fixed
prior to an actual event.
-
Expandable to whatever size required, within limits.
-
Easy for exercise ''referees" to view activity without being intrusive to action.
-
Limited training required to learn how to use software.
-
Generates interest and enthusiasm from the next generation of analysts in tackling
today's attribution challenges.
(U) Some of the possible disadvantages of this approach may be:
## S~Rn
-
People might not feel comfortable adopting the new technology.
-
Less than perfect technology may not yet be immersive enough to make it seem really
"real."
-
There would be some IT and accreditation costs to host commercial software on
classified networks.
-
Participation may look like "goofmg off'' to certain managers who do not understand
the purpose of the effort.
(U) Collaboration tools, such as wilds, could also be used for ''virtual exercises" and could greatly reduce overhead and costs. No one would have to go TDY and people could fit their
"moves'1 in when they had time available. The wiki could simply be a clearinghouse for
"game" moves. The idea of a virtual exercise would cause far less disruptionjp. W.,.y-to-day activities and would allow more frequent tests ofthe !I
process, costing ot!i~:~hat it would take to form a small group of people to design ~mutation with realisii'C;;~;;..
-
information. These people would not even need to be physically locatec:H~g!=.ther."":E~~rtises could be designed on a wild over a period of time by allowing expe~ yi'criiiw.Jt>,pte _to ihe exercise as time permits. What would be very interesting ~~;~o use ~~~i~~ inchia~::a;"red team" in the exercise who could respond to the IC's moves:Th~y coum::~~ge their actions depending on what the "blue team" did, providing for a more ~e~Jt~tic sceiijio instead of a planned out one.
.;::::.
:;;~:;::..
. .: -
)~~=} ~
~-:~~~;;~"
~~=~====~=:~~-~
:--~~.;; .. :-.
.
(U) Competition is Good
.. -- :::. .
::;::;~..
.
(U) Including competitive aspects to vi;1i~~~~~!ci~~~: will a~~~::rate creation and sustainment of a secure on-line attribution co~\Jpity. Analogous to "fantasy baseball,"
virtu~l attribution "teams" comprising individual~~~rii different agencies, offices, and programs and would compete against eac~:i>ther 9\:iddressing attribution-relevant topics.
Each virtual team could have a !I!~~ o(!Nf', IC, and LE expertise rather than only one discipline (e.g., a LE-only.m~mS'ers team). The composite mix of the teams coupled with the competitive environmenf wpuid fo~ter and accelerate 1) lexicon and semantics development between ~q ~!lg ~
(see (U) ORGANICALLY GROW A LEXICON), 2)
sharpen members' a9~~yticafii(~~Yaft with sustained exposure to different perspectives, 3)'
foster coll!<&f~J~!!Dq;te~work habits that would morph into effective best practices during actual n.tJCI"eai;~~butioti:~"Y~t\ts .
.
/:{~::;;,_ _,}~;
':~~::;
(U) Outlci~l(.
, ~::~: ....
(U) Success ~~~~~tiding the ~f the future will require cooperation from all organizations involved and a commitmen~owing and maintaining the next generation of nuclear experts needed to take the L_jnto the future. For each recommendation chosen for implementation, we recommend a high level "Champion" be identified to shepherd the process into reality.
## Recommendations
-
The USG must start building a stronger nuclear community, not just as it applies to
the attribution problem, in order to ensure sustainmept of current capabilities into the
future.
## (U) Suggestions: Required Resources
(~F) There are a::{k fun~~~~~~ assumptions that can frame the debate.
## (U) Appendix A: Index Of Recommendations And Findings
(U) Recommendations (Not in any order)
(U~buOl'~------------------------------~1
(U) Applying Catalyst to Nuclear Attribution . ='
;:=:;:;::..
)'
An information technology (IT) solution- .b~&i;<in:tlie Catalyst program- is recommended to allow a single "switch" to be thrown Ui~iwill gf~~ all rele~~t players in the emergency the access to information and informacl(;i{ifrols hl~r she neetis. Based on a subset of the Catalyst concept, the architecture and moda.'iii1~~;{or interactive information sharing and analytic exchange across the IC communities c6uld:b~readily created to improve the quality and speed of the analytical work dealing ~ith a c~ihle WMD threat or event. In addition to the obvious agility and potential for syq_~tgy, it is'expected that this capability could reduce the assessment time and imprpy~=~~cuf.icy more than almost any other approach: While the Catalyst Project is focused gri tl!e'intel!igence Community, the concept of a fully interactive community with conc~en(~warene~; connectivity, and access to appropriate elements of the total database can tjfteaallf.extenaed to include law enforcement and technical forensic elements. PJ:wning,i~fully ~1:fp;~viding appropriate permissions prior to an event can allow the;riipia::fQnniiig~q:f...a ll!fge virtual working group that can much more quickly intem:;l\t~ and ~tiluate di~;iift(i can facilitate rapid overall synthesis and suggestions for additici@..~olJ6Stion and afualysis. [(U) INFORMATION SHARING AMONG
DISPA.Ji:XTE KINGDOMS]
~~~:;~~;~ ..
Mi I
## (U) Using New Internet And Media Technologies To Enhance Information Sharing
1
~hould ensure that the channels for sharing information and expertise that will be requ1red to make a credible attribution assessment are in place and regularly exercised so that they can be used effectively and immediately in the event of a nuclear event The use of new Internet technologies, e.g. Web 2.0, and new media technologies should be strongly considered as a means for achieving this goal. [(U) INFORMATION SHARING AMONG
DISPARATE KINGDOMS]
(U) The Technical Event Manager (TEM)
..~.:- .
-:~:~;~;;~:.,~;;:
Access to all-source information early in the analytic process may undql~~bi~;tJte u;a-ifuical analysis. This risk can be mitigated, and the benefits of inf9Jillation;~h~ng ili~p._q.ined, by providing event-related all-source information to the Techiii~al ~vent=~~ager.{TEM). This will maintain the benefits enumerated in the planning and intem~~tion''6~~e"technical data and provides context for the technical analyses. *[(U)* Limiting"Bi~::irt,Techiiical Data Interpretations]
.. :?
-:::~;;;:;::
(U) Nuclear Community.)0~-9/
_;;~
Start a nuclear anal~t;social:gr~up;:~here all nuclear analysts from the intelligence, law enforcemen!,,.~d n~~!~ foreiisi~''~ommunities can participate. This would not be another workin~_gioui?.i;:~~t a 'sli~J?.~rt !letwork intended to bridge the expertise gap within the nuclear comn:t\JPity. Tli!~ would-b~)goographically limited, but "chapters" in different parts of the counicy::t:,ouldfoim and begin to socialize. Augment the traditional social groups with the use of emergii}:g!~~cial networking technology to create a "virtual community" of nuclear experts who hold r~~l!!f, ~formal chat sessions on the appropriate computer systems. The only limitation in tlli~;!icenario would be access to appropriate classified computer systems. Most analysts have .access to at least SECRET level computer systems, and many have access to JWICS. Establish a formal nuclear mentorship program, where junior analysts are paired with nuclear experts both inside and outside the US Government. Modify or adapt commercial software to allow nuclear analysts a new way of participating in exercises related to nuclear attribution. Consider the use of programs such as Second Life or a modified "first person shooter" video game as a platform for a scripted exercise. Use social networking tools as a means to create virtual exercises with adaptable adversaries to test thec=::process for attribution. [(U) BUILDING THE NUCLEAR ATTRIBUTIO~ COMMUNITY 2030]
(U) Crowd Sourcing To consider the benefits of"crowd sourcing," test the possibility of expanding the collective beyond cleared and individually authorized people to the larger community in the unclassified environment to evaluate the benefit of engaging this huge resource. [(U) THE
RIGHT STUFF: HARVESTING EXPERTISE]
(U) Characterizing TNF Capabilities and Limitations
(U) An existing limitation of interpreting TNF data is due to an incomplete knowledge of foreign nuclear materials stocks, foreign nuclear weapons design, and an infinite design trade-space for IND designs. In addition to studying the capabilities and limitations of producing TNF data, a separate study should be conducted- possibly under the auspices of the JAEIC- to characterize our posture to interpret TNF in the context of what is known, and what is not known about nuclear materials and designs. Such a study should i.dentify means to reduce or eliminate these deficiencies. [(U) HOW TO **BEST USE TNF** '::;;;/'"
INFORMATION]
}::;.t..
## (U) Findings
On the .topic of Lexicons and Communication:
- (U)I
- Ml
- N1
-
.
-
(U) Use terminology universally recognized an<!:~~~epted ~~~{i~ommunicating
internally and externally. [(U) SPLI'!~~~pND:~P,.f;,CISION-MAKING: A LAW
ENFORCEMENT PERSPEC~)'
-:~:::~~:
';:~:i;;;.;.-'
-
(U) Prevent the "CSI Effect" by.c:9~uni~iing and ~ducating the customer. [(U)
SPLIT SECOND DEC1SION-MA~G: A LAW ENFORCEMENT
PERSPECTIVE]
:::;::~. .
.
-
(U) Speed of technical nuclear analysis ci'~:orily proceed as fast as the laws of physics
allow. [(U) SPLIT SECOND D:tciiSI()N:::MAKING: A LAW ENFORCEMENT
PERSPECTIVE]
.. ::.. ./i::'
-
(U) Differences in cultifre;:C!Iifl ~pply additional pressure to a pressurized situation,
which can spin "Wh~t po:we kitow?'' and "How do we know it?" out of controL [(U)
:~~~~~i~ rii;~~~~g~~MAKING: A LAW ENFORCEMENT
-
(U);M~~ irif~~~d intelligence-driven decisions, by find multiple sources of
. .;infoiiii.~~~n to s\lpP.Qfl: exiting data and .to capture differing perspectives. [(U) SPLIT
-
(U) Since the nuclear attribution primary communities are small and closed, much
more frequent interactions among the communities is recommended t<t.~r~te the
similar effect of developing a common lexicon with many interactions:~i9i~ massive
social network. [(U) ORGANICALLY GROW A LEXICON]
. ;:::~;.
,
-
(U) Promulgate a mechanism for ongoing communication and i~t~-~on, 'ifi1~dltion
to the mini-exercises, to further facilitate development of a c~nyli~iit~~j9on
indicative of a growing community. Virtual means;,<?,[ inte~~#S!i may b~"!>est suited,
as community members are geographically disperseo. [(U)"OR:?~CALL Y
GROW A LEXICON]
.-------'';..;.;::!i.l::
;~l:/
-
(U) Information Technology (IT) is integral!
l~ft"c?.tt, as if facilitates
communication and ultimately enables effective,I;Pid timely ~itlysis. In order to
leverage the benefits of the IT availab,I~j
: ~,it will beimportant to plan and
coordinate these tools before the s~ofaqj~~ercise:~f.;~actual crisis. [(U)
INFORMATION SHARING ~9NG J?-jSPARAtE' KINGDOMS]
-
(U) r--1oformation sharing ~houi(J::t,!lkt(place within a common computer system
inte~cleared to the highest practi~=Jc;vels. However, given proprietary data
among organizations, the NCTC should.iii!,Y,~,a key to databases with assigned points
of contact, through which aceess ~q:~divi~iial databases may be enabled. The chosen
II
computer infrastni~ .. W!e ~~gfiid be configured as soon as possible, and all c=J
'riieiiiliers should be Jl,pi'fti~!:Yet thoroughly familiarized, trained, and exercised on
~ese systems, appli~~~Qns, an(! protocols. [(U) INFORMATION SHARING
AMONG DISP.AR.t\:;{E KINGDOMS]
On the topic of Collaboration:
-
(U) Collaboration teams sho~ld be identified and begin working together to establish
effective working relationships prior to an actual nuclear related event. [(U)
INFORMATION SHARING AMONG DISPARATE KINGDOMS]
-
(U) It also is recommended that collaboration should not extend down to the those
involved with basic generation of data in order to prevent initial bias or cause data
generation to be diverted away from what may be eventually proved to be valuable
information. [(U) INFORMATION SHARING AMONGDISPARATE
KINGDOMS]
-
(U) Have working groups meet often and practice not only on exercises but real world
events.
THE RIGHT STUFF: HARVESTING EXPERTISE
On the topip;.of;:rraiiii"g.and Ex~~cises:
':::::@ Thti:increased:frequency interactions among the communities could take the form
6~P.t!lritltly or bi-monthly secure, web-based, virtually-linked mini-exercises on a
parti~\llar facet of the attribution process to calibrate the communities with the
spedfi~::wnns and their meaning. [(U) ORGANICALLY GROW A LEXICON]
-
(U) T~~-inini-exercises should be brief (e.g., -2 hours) to foster and stage
development and acceptance of a common lexicon and to minimize disruptions to
ongoing mission areas. [(U) ORGANICALLY GROW A LEXICON] ]
-
(U) Explicitly train c=:Jmembers on issues of biases, heuristics, intuition, bounded
rationality, etc. [(U) THE ROAD TO HELL IS PAVED WITH NORMATIVE
COGNITIONS]
-
(U) Members ofthec==]should be selected based upon criteria that include the
ability to be aware of, and counterbalance, the influence of heuristics and internal
biases. [(U) THE ROAD TO HELL IS PAVED WITH NORMATIVE
COGNITIONS]
-
(U) Include elements and activities invoking biases and cognitive mind traps that
undercut attribution analysis into ormal exercises- (to be included in quarterly
simulations or other exercises forD[(U) THE ROAD TO HELL IS PAVED
WITH NORMATIVE COGNITIONS]
On the topic of A voiding Analytic Minefields
-
Do not lose objectivity by deviating from the establishe(j investigative plan. [(U)
SPLIT SECOND DECISION-MAKING: A LAW ENFORCEMENT
PERSPECTIVE]
-
(U) Lower level employees should be empowered to make certain decisions and be
held accountable for them or without concern for negative consequences/punishment.
[(U) SPLIT SECOND DECISION-MAKING: A LAW ENFORC~~~T
PERSPECTIVE]
.
:;~;:;::
-:,.:!-: ... '!'
-
(U) Build into the decision-making process critical review by collea~es -~l~::.A
members of other groups, as well as other checks and balances,:i6.-;niiig~te nis'h.to
judgment (bounded willpower). [(U) THE ROAD T~ HE~t~:!$ PA ~~::WITH
NORMATIVE COGNITIONS]
:.:
.
' ::::~::..
. ~
-
(U) Cr~ate s~ctures and processes to enablerl~~~f~~~~~::~~:lf~to wh~te~er
degree IS poss1ble) from the effects of extrem~ot co891fion) on 1ts attnbution
efforts. [(U) THE ROAD TO HELL.lS:;~;A VE~D;~WITH N6RMATIVE
COGNITIONS]
.: .:.:.;.;..
. .;.;;:..
.
_,:::~~;-
~=~~:
"'-!:~;~~.~;
-
(U) Th~
lhould develop fu~~\!~.i~gate maladaptive group dynamics. [(U)
PERILS AND PITFALLS OF GROW!~]
-
(U) Thec:J;hould avoid over worry iih!>:iWgroupthink, but avoid insularity of
perspectives, consider membershiP.:~riteria=;io create balance within group (across
status, expertise, etc.), aq~_recQgpl~e problems associated with new groups. [(U)
PERILS AND PITF'~~$";0fGROUPSJ
-
(U) The c=:J shoulg ~'Ogniz~ the need for inoculation of personal and information
networks prior J~;ev~~ ~4:w'eet regularly for joint exercises to build the
relationship~~.J.~perience~~~ssary before a real crisis. [(U) PERILS AND
Pllf:~L$~QJ.i!.pROUPSJ
- . .Jtrj''n~O sli~~!~,p~es~rve spli_ts" invo.lving differences of opinion i~ re~orting to
.:::~:::~pnsull!~ ana eny>haslze anformabon shanng and enhancedc::::Jcoordanabon. [(U)
.
::e~RJI;S AND PITFALLS OF GROUPS]
-
(ti}:~.~ CJ should counter the debilitating effects of a high stress environment on
C=JB.~~niploying stress monitors. [(U) PERILS AND PITFALLS OF GROUPS]
-
(U) Th_e'l=J should select a mixture of experts with differing competencies (both
task and general) for c=:Fbgroups to enhance flexibility in dealing with
ambiguous environments and coordination across the three communities. [(U)
PERILS AND PITFALLS OF GROUPS]
## Best Practices Recommended ~ I
-
(U) Assimilating, analyzing, consolidating, summarizing, and reporting nuclear
forensics information must be a two-pronged iterative process- continua.lly building
the case and succinctly reporting findings. Employ dynamic analysis methods that
weigh hypotheses against evidence. As time permits, use data modeling and
visualization tools to gain deeper insights and accentuate reporting. [(U) CAN
OTHERS UNDERSTAND YOUR DATA?]
-
(U) Leveraging metadata successes implemented by other communities. Use of
metadata would facilitate and accelerate effective sharing of data and information
among the nuclear attribution communities as designed and developed to meet each
community's needs. [(U) Using Metadata to Foster Information Sharing Among
Separate Communities]
-
The Administration might r~uest authorization/appropriation of anc=J>perational
fund, to be available if/when_
~s formally mobilized during a Level One or
Two Threat. [(U) SUGGESTIONS: REQUIRED RESOURCES]
-
(U) The c=J:eeds to recognize that pressure for immediate responses will be
directly proportional to the magnitude of the consequences surrounding an e:vent.
Through exercises, the c==lhould educate senior consumers about the reality that
technical analysis, crime scene investigations, or IC efforts are likely to lag behind
their more rapid desire for actionable information. [(U) BRINGING ~RIGHT
BROOMSTICK TO THE WIZARD]
:- ;::::;:..
-
-
(U) The c::::::J needs to recognize how consequences affect risk ~g!~.l)ce of:R9];icymakers. [(U) BRINGING THE RIGHT BROOMSTICK ~9.~niEi:}.Y,!ZARD]
-
(U) ~ell must calibrate co~uni~tions with.i~~nior c~R.~ers b~~d'upon the
way dttterefitleaders structure thetr advtsory systems [(Y} BRIN.~P'li9THE
-
.~
,.,.., ~ ~ 4
RIGHT BROOMSTICK TO THE WIZARD]
;:~;;:;~,,.
:::
-
(U) Thec=Jneeds to understand that people res~rid to'~'S~W?!:l~J:s'~frisk based upon
whether these are presented in the domain of gaiAs::or losses.l(V) BRINGING THE
RIGHT BROOMSTICK TO THE_,"WJ~~j;:;:~:::~..
.-
-
(U) The r--lneeds to recognize t}}it if ati;pJjution iitf9f!liation being presented to
senior p~makers diverges fOO~;~eir P.(~:.Cxisting?deeply held b~liefs, more
evidence will be required to gain tfie~;~~ehtion [(U) BRINGING THE RIGHT
BROOMSTICK TO THE WIZARD[:~;:~
.
-
(U) For the II
to get the right answer,:~~ey 'must learn how to ask the right
questions. Deallng with the factQ~:around'information seeking will be critical to the
c:=::Jand its supportingJ~:x~rs.;:f.liey shoUld consider: Education on the information
seeking process, to a~oi'd';$rrupon pitfalls and overcome natural frustrations, and be
able to optimize theif Qwn beh~viors; creating social norms that accept imperfect
knowledge am~~g ni~m~~~~;:l!fid reward outreach to acquire the right knowledge and
expertise; g~W.yg into'~tli~;~:mind of the client" and incorporating interactive strategies
to ~e.ep;Jpformation seeking focused. [(U) BRINGING THE RIGHT
.JJROO~TicKi:'!(}.-:THE WIZARD]
.~;:::;:J:he c::J should i15e the "right" kind of expert advisers to interact with policy-
'
;~-~-"[(U) BRINGING THE RIGHT BROOMSTICK TO THE WIZARD]
-
M:ftt:qmote a healthy culture: Thec==:J as a new organization, can seize the
oppo~f;y to explicitly shape an organizational culture designed to optimize
functiqnlng and mitigate or eliminate the effects of known negative factors. This
culture could include behavioral norms and attitudes such as: open communication
and information sharing, minimal in-group status distinctions, inclusiveness
(permeable boundaries for belonging, especially with regard to analytic aspects, trust
in intentions, common group identity, shared sense of goals and mission, role of
devil's advocate or red teaming, understanding the strengths and limitations of
expertise (link to expertise section), and healthy interactive strategies, such as
negotiation (link to negotiation section). Consultation from occupational health
professionals should be obtained regarding mitigation of the physiological aspects of
stress. [(U) PERILS AND PITFALLS OF GROUPS]
-
(U) Maintain and grow social networks: Well-connected and well-structured social
networks, among their many benefits, improve performance in groups facing complex
problems. c=]members must maintain and leverage their existing networks while building new relationships both within and across group boundaries. Regular, formal assessments on network structure and individual position and performance within the network, using methods from social network analysis, will provide a means to evaluate the current functional status and suggest areas for improvement. [ (U)
PERILS AND PITFALLS OF GROUPS]
.
-
(U) The rl
needs to manage expertise effectively. Expertise, while often critical,
can be a doii6\e-edged sword in groups, genr 1ic4 rigidity, and excessive
deference. To leverage expertise effectively, the
should consider:
o (U) 1. Selecting for fluid expertise in its mem rs, and encouraging and
teaching fluid expertise methods to the group.
o (U) 2. Incentivizing collaborative, instead of allstar behavior.,~ ExP.ertise alone
is not sufficient. Metrics for individual success should include!~-~perative
behavior and information sharing. Provide actionable feedbacK' icl'!iJ.ldivic:juals,
such as a personal social network analysis. .
.-::>
=t~~,,;::
o (U) 3. Building ~decentralized social networks fodiittif.ination si'i&ing
across the entire L_j Measure and evalua!~. these 9fg~izati6iiitJ::social
.
..
.~.....
..
;!! .. ::;-:-.:.~
,,.~
networks, and take steps to Improve them asdeficiencies::are observed. Share
results.
.:::;:...
-:~:::;-:.
o (U) 4. Institutionalizing these practices and.metrl'~~it~,.sha~'ilie culture. [(U)
PERILS AND PITFALLS OF GROu:rsf'
::::;::
## (U) Appendix B: Attribution Terms A
(U) Accelerator Mass Spectrometry (AMS) - A mass spectrometry techilique that uses a high-energy particle accelerator to measure rare isotopes. The technique was originally developed to measure 14C in small samples. Recently, AMS has been applied to measuring uranium and plutonium. Like other forms of mass spectrometry, AMS requires extensive chemistry to dissolve the sample and to separate and purify the element for aq~ly~is. AMS
instrumentation is large, complex, and expensive, and is found in only a smalf~q(itnber of laboratories around the world. AMS excels at measuring exceedingly small isoiiift~-~ati<?s
(e.g., AMS can measure the 14C/12C ratio to as small as lx 1 0" 15). AMS C!!Jt,p~tediij{t:e;W as
1,000,000 atoms of plutonium, and can measure isotope ratios to bette~;dJ'iliii~~?.p~-in:=ioo precision for larger samples.
::;;;;;
:~:=h -~
:.~:!:
" - .. .
~ ~ p
.' ~ -~:t=!~~~-
.;;
(U) Activity (or Radioactive Activity)- The decay rate ofth~;rMioactl~~;fi~topes present in a sample. The activity is measured by detecting the alpha,p~ici~H~~taparticles, or gamma rays emitted by decaying isotopes in the sample. Activi~~l's"measu~;in units of decays per time, including Bequerel, Curie, and dpm. 'fl!~-:t~~ ",sp~~f}c activitY" refers to the activity per mass of sample, element, or isotope. ~-~e'also:~ta parli4l~ ... ~pha particle, and gamma ray.
.-:~::t~h:._,:)f
::;:~: ..
(U) Actinide Element - A group of elements~P,.~;t!te periodic table of elements that includes uranium, plutonium, thorium, neptunium, amerl~i~.and curium. These elements are chemically similar and have been groupe<twith th~'ilightest element of the series, actinium
(hence the name actinides). The actinid.~S;have elements with high enough atomic weights that fission becomes significapt;:::1Ibe a~tinides are also chemically similar to another group of elements, the lanthanide ~fe~eiits (a)so called the rare-earth elements).
~= ....
=~
(U) Activation Ana~y~j~i~ ~=~~~AAi.qJ~ for measuring trace quantities of elements iri. material by making th~P.l rad.i:9J(ctive by neutron irradiation. The result is an elemental signature, or fingerprj!lt{~!:~ be:~~ tQ determine the elemental composition of the sample.
Acti':~.~i~n anal5fsis, or n~~!fo'n activation analysis, is often used in forensic investigations.
Theiio.i9~.ctiyjcy of the activated sample is usually negligible. Activation analysis requires irradiati6ii~ii{a reactor, radiochemistry to separate each element for analysis (although some elements ~!}:.~~tected without radiochemical separation), and radiation detection to measure the a.tD.olint of radioactivity produced by the activation. Both beta particle and gamma ray detection are used for analysis.
(U) Activation Products- Generally, radioactive isotopes produced by neutron irradiation of a material. An example is the production of radioactive cobalt (60Co) by irradiation of stable cobalt e9co).
(U) AEA - See alpha energy analysis. (U) Age-dating - A process of measuring the amount of decay of a radioactive isotope in a material to determine the length of time between the time of imalysis and when the material was last chemically altered. The amount of decay is determined by measuring the decay
product (or daughter isotopes) of the radioactive isotope (often called the parent isotope).
Both mass spectrometry and radiation detection methods are used to measure the amount of parent and daughter isotopes in a sample. An example: measuring the parent isotope 234U and the daughter isotope 23'7h to age date uranium.
(U) Airborne Radioactivity - any radioactive material suspended in the atmosphere.
(U) Alpha Particle Radiation- A positively charged particle made up of two neutrons and two protons, emitted during the radioactive decay of certain radioactive isotopes. Alpha particles emitted by the decay of radioactive isotopes, which include most uranium and plutonium isotopes, have specific energies that are characteristic to the specific isotope and decay scheme. Alpha particles travel approximately one inch in air and can b.~ st9:pped by thin layers of light material such as a sheet of paper. While alpha particles po$~:-iio direct or external radiation threat, they can pose a serious health threat if ingested or inli'iii~~.
~
........ :...
~:~:~~=to .. -,:;
(U) Alpha Energy Analysis (AEA) - a radiation detection technique tl:!iit;ffi.~pres tiii
energies of alpha particles. Alpha particles emitted by the d~ay of.r:j,4{oacti~~:;l9topes have specific energies that are characteristic to that isotope and d~cay scl:ieift~;:.AEA.ean be used to identify specific radioactive isotopes and to determine the aW9H;_nt (or"a~ij~ity) of that isotope in a sample. AEA normally requires chemical dissohiti6ii~~sl purification of the sample. The detector types include solid-state detecto~ . ga$""propt)fi\q~~il detectors, liquid scintillation detectors, and solid scintillation d~~~~rs}~.t\ can det&t quantities of radioactive isotopes as small as 0.001 dpm,...:-~.:;:.:;;:~:;;.
~~;::;:-.
- .
.
'~;::::~
:::?
~~::~:-
(U) Alpha Spectrometry - Another nam~::(~r:Npiia Energy :A.nalysis or AEA.
~ ... ....
-:-::".
(U) AMS - See Accelerator Mass SpectrometrY;:~:~~~~;::
(U) Analysis Class - A term that describeithe lev~l of effort applied to the forensic sample and the degree to which the for.ens'ic rui~-can be interpreted. The analysis classes are ordered in the sequence of their exe~tiqri:'det(!ction, identification, characterization, and attribution.
That is, an event happe~, s9.~~thingj~ detected, samples are taken, and material is identified, the materiaJ.:is'chiii'ii<;t,eriie<l and, if needed, the forensic data are combined with other information fof.:~itributiort;:-.-
.. -:;::!::::...
..~ ~::::.: ...
(U) ~~9~~~ ~~~~rpti~'~::~pectrometer (AAS)- A technique that measures light absorption to idefitify and~C:letermine.the amount or concentration of an element in a sample.
..... ....
."&
~-::: ".
(U) A.to~i~=~~P_<!r..Laser Isotope Sef.aration (AVLIS)- A VLIS uranium enrichment technology is"'ti~ed on the fact that 23 U atoms and 238U atoms absorb light of different frequencies. Although the absorption frequencies of these two isotopes differ only by~ very small amount (about one part in a million), dye lasers can be tuned so that only the 235U
atoms absorb the laser light. As the mu atom absorbs the light, its electrons are excited to a higher energy state. With the absorption of sufficient energy, a 235U atom will eject an electron and become a positively charged ion. The mu ions may then be deflected by an electrostatiC? field to a product collector. The 23Bu remain neutral and pass through the product collector section and are deposited on a tails collector. Although A VLIS technology appears promising, it has proven to be extremely difficult to master and may be beyond the reach of even technically advanced states.
(U) Attribution - Nuclear attribution is the assignment of responsibility for the intended or actual use of nuclear or radiological materials/devices in criminal acts or acts that threaten national security.
~/
:::::: ..
(U) Autoradiography - The process of making a photographic image or map of the radioactivity of a surface. This technique is widely used in biochemistry and genetic studies;
it is also used to study oil paintings. The radioactivity of the activated surface involved is usually very low. Autoradiography is extremely useful for locating radioactive particles
(sometimes called "hot particles") in samples, so that the particles can be isolated and analyzed individually. Autoradiography offers an alternative to fission track analysis for particle location and identification.
B
.-;::. , .
::;:::~==)
(U) Background Radiation - Radiation from natural and man-made SQ.iftces~;:!~cludifig from natural radioactivity in the environment, cosmic rays, and f!-9m radis~~!ivity rel~ed from nuclear power plants, nuclear production activities, and nuclear weap~;~sting::'
.:~:=:~~
... :::~:~:-~
(U) Bac:ksc:Btter - A method of identifying material or m~prlng;~p_ating.tfllckness by measuring the amount of radiation reflected back from ~.Jieiiln proj~e~'into the material.
~ .. ~.~
~;;!~::::.~
.. :~
(U) Becquerel (Bq) - A measure of the inten~it;~(htdio'icitiv.ity,. One Becquerel is one decay (or disintegration) per second of ~.ri@.ioactiY,~=isotope:==~i::3q equals: 2. 70 x 10"11 Ci or
60 dpm.
/::::;:;.. ./
(U) Beryllium (Be)- A chemical element ~J~~:~.veral useful nuclear properties, including a high cross-section for reflecting neutro~ ;~Ptd a "Iii@t cross section for absorbing alpha particles and producing neutrons. Becy,lJirlfu metill can be used as a neutron reflector in nuclear weapons. When exp9~~::!9a ~ifong alpha emitter (such as 21'1>o), beryllium metal will generate neutrons, whi~h c:w"initi~te a chain reaction in a critical mass of fissile material.
.. :
.~ ..
: ..
(U) Beta Detector - ~~=o/~i~!=~~::aes of radiation detectors that detect and measure beta particles emitt~d fn;.Wradioactfve'isotopes. The detector types include solid-state detectors, gas proPQrtfo~t:.9eie"&Ats, Geiger counters, liquid scintillation detectors, and solid scint~l,l~tion d~(~tors. fii:~:most sensitive beta detectors can detect activities of 0.1 dpm or le~s::::::::.,.
./'
,.
.
:~::~:~ .... ; ....
(U) Beta -~~):uc:le Radiation - An electron that has been emitted from a nucleus. A beta particle typi~itY,;:li~s an energy between 0.001 and 4 MeV. Its penetration in material is short, but nucli<ies deposited on the skin can be an external radiation hazard. Beta emitters deposited within the body are a serious internal radiation hazard.
(U) Bismuth (Bi) - A chemical element that can be irradiated by neutrons to form 21'1>o, which can be used in neutron initiators for nuclear weapons.
(U) Boosting - The use of a DT reaction to produce 14 MeV neutrons to enhance the fission in a nuclear weapon. Boosting can greatly enhance the efficiency of a fission weapon.
(U) Boron (B) - A chemical element that has a high cross-section for absorbing neutrons.
Boron is used as a neutron absorber in control rods to control nuclear reactors.
## Se~Rn S~
(U) Bulk Analysis - The analysis of macroscopic amounts of material to determine the concentrations of elements and isotopes in that sample.
(U) Burn up - A term used to characterize the extent of use of nuclear reactor fuel. Bum up is a measure of how much of the fissile isotope (usually 235U) has been consumed in nuclear fuel. Typical units are megawatt days per ton of uranium fuel.
c
:; ..... ,
.:;::~!:: ....
(U) Cadmium (Cd) - A chemical element that has a high cross section ~QJ;.~psorhi~g:~-;)
neutrons. Cadmium is used as a neutron absorber in control rods to ~~~~6"l"I\~~~~-r~actors.
(U) Carbon Composite - A material consisting of carbon fiBer "clciffii~~impre~ii~d with
epoxy resin. Carbon composites have high strength and light W,~jght, an"tirprut;"easily be shaped prior to the application and setting of the epoxy. The.iDe'cfilWicallilid manufacturing properties are highly useful for gas centrifuge manufac~~?-Carb~&'~niposites also are used extensively in aircraft and missiles.
. . .::..
,.;~;;::;:.,
.:
.
1"-"."':..~ . .. ~ .....
(U) Centrifuge- A rotating vessel that<;;~:,. u~~~}or the-~t~h~ent of uranium. A device that applies centripetal force to a sampie;~~~ly:Via motor-diiven rotary motion of the sample. There are many different kinds of e~mfuges, often for specialized purposes in research laboratories, medical labs, and material:i~~~h. (See also gas centrifuges).
.
-=~
.... :::" ..
(U) Ceramic- A hard, pottery-like ma~~6iii with a high resistance to heat (e.g., oxides or carbides of metals). Fuels fo~:Pil.~!~i'~actors operating at high temperatures are often ceramics (e.g., ura!lium dio~idc::, .. uraniym carbide).
:"~ .-:.
.~:
(U) Characterization;~~~aegfl~catiofi::of material and device properties. This information is vital to perf<J~ing .~i!tJJJution, "6~i""ii also may be sufficient in and of itself to perform
-
-
~~'
i .:. ..
attnbutt9~~=~::;:;t;!~.
:::~:::: ..
. ;:~r
~::::
::!:::~;;
(U)-<::ii~~ical/Assay- Ch~mical titration and controlled potential coulometry are standard methods;r9~~~etermination of the element concentrations of uranium, plutonium, neptunium or other nUij~f:~~9.~Ponents of nuclear fuel material for accountability meas~ments or accountabilitY"Y.efifications. In chemical titration, the sample is made to react with an exactly measured amoiint of a selective reagent of known composition, leading to the completion or characteristic end point of a well known stoichiometric reaction. Titration methods are designated according to the mode of detection and the end points. In controlled potential coulometry, the element to be analyzed is selectively oxidized or reduced at a metallic electrode maintained at a suitably selected potential. The number of electrons used in the oxidation or reduction is a measure of the amount of element present in the sample. The precision and accuracy of these methods is better than 0.1 %. They are well established and used routinely in nuclear accountancy and safeguards laboratories. They therefore can be very effective for the characterization of interdicted material, provided that samples of at least a few tenths of a gram can be made available.
(U) Contamination - The deposition of unwanted material of any type (radioactive or chemical) on the surfaces of structures, objects, or personnel.
(U) Conventional Forensics - The application of forensic science to conventional criminal cases. More specifically, conventional forensics is the scientific analysis of samples, things, and people with the goal oflinking places, people, things, and events. Conventional forensics includes the analysis of fingerprints, DNA, fibers, hairs, paint chips, pollen, etc. See also Forensic Science.
{U) Counter (or Radiation Counter)- A radiation measurement system that reads out the counts or countrate directly, in contrast to a dosimeter, which reads out in units of radiation
~
~-
.
~~;::~~-
(U) Critical Mass, Critical, Criticality - The smallest mass of fissile material"th'~-~ill ~low a self-sustaining nuclear chain reaction, or criticality. At criticality, the a.~.9:n>tion~ij.~;9f neutrons is just balanced out by the production of new neutrons by fissi9ii:';Th~. criticli:f mass depends on the type of fissile isotope, its chemical form, ge_Qmetricat:-rangerrt~t, . and density; the amount of material needed is decreased if the rr.iiiterial i~"b~@.P.resse,d'!' Nuclear power reactors need to operate at criticality to produce electrici&. Nuct~;weapons are based on highly supercritical designs.
.,- ' -::::.11:;:-,
.. :-:
~--=~
......
, .. -.
'" .. :~~~-~ .. ~.
(U) Curie (Ci) -A unit of radioactive decay ~~.j~ .. ba!,!~4;~~ the acti*iiY of one gram of radium. 1 Ci equals 3.7 X 1010 Bq or decaysp~t~~prid or'~~22 X 1012 dpm.
,.;~~:{;:;.. )?
')~;;-
.
~~:=:=:-~y
D
.. -~==~~j_~.~~--~~:,/
. ,. ....
;::====:
(U) Daughter Isotope - A Jlaj~~~~i~e;,isotope (called the parent isotope) decays into another isotope which is .called .fu~ d.a:uibter i~9tope or decay product The daughter isotope can be either stable or radio,;t~Hve, d~9!~g on the parent isotope.
(U) Dec~y~~iti$,~~~~~~~.~us .=formation of one nuclide into a different nuclide: Decay may inyolve tnetemission~~ofalpha particles, beta particles, positrons, or gamma rays from the nuct6\isbbe ca~b.tre of elegtrons by the nuclear, or fission of the nucleus. A decay process is charact~~~(by a half-life (i.e. the time for half of the atoms of a radioisotope to undergo decay). Ai<>~ealled "radioactive disintegration."
~-:~:~:,;. ... .;!'
-.-.. ~
(U) Decay Pr.iduct- The outcome of radioactive decay. See daughter isotope.
(U) DeUberate Signature - A signature, such as the isotopic composition of HEO that is controlled and specified. Deliberate signatures are essentially product specifications.
(U) Depleted Uranium, D-38,or DU- Uranium with a concentration of 235U smaller than that found in nature (0.72% atom%). It is largely obtained as a by-product or "tails" of the uranium enrichment process or obtained from spent (used) fuel elements. This material has low specific radioactivity and poses no significant risk to human health. It often is found in aircraft as counterweights and in boats as ballast material. It is also used in anti-tank or armor-piercing ammunition to enhance penetration.
(U) Detonator - A device that initiates the detonation of a charge of high explosive by subjecting it to percussion by a shock wave.
(U) Deuterium (D) -An isotope of hydrogen in which the nucleus has one proton and one neutron (nonnal hydrogen has no neutrons). Deuterium is useful as a neutron moderator (in the form of heavy water) and has nuclear properties that are useful for thermonuclear or fusion reactions.
(U) Discovery Class - A term that describes at what stage a nuclear or radiological event was discovered. The underlying assumption is that the unfolding of an event consists of a sequence of processes that lead from the planning to the execution of an even~,.
._
:ii~,.~.::
(U) Dirty Bomb - An explosive device that is intended to spread radioactive ~$!i~l fr~m the detonation of conventional explosive . See also "Radiological Dispe~!!lJ.Jevice.;~~D)."
~~=~ ~-~-::;;~~=~..
~- ..
(U) Dose- A general term for the amount of radiation abs~IY.ed ove~ll;~period'~fj~~iJ;ne.
~:~~.
.~ ~!~~::~....
!~-
(U) dpm (decays per minute)- A convenient unit ofradioactiY":~-decaY::t:j;;dpfu equals
13
I! -~~ -
4!1Jr
0.0167 Bq or 4.50 x 10 Ci.
..:-;/
.:~:;::..
..-.
.
.: .. ~--.~
':~~~;~~~~;
(U) DT Reaction - A thermonuclear or fusion r~tion:~(\yeen the J:lydrogen isotopes deuterium (D) and tritium <n that produc~~;-t4'N1~~,ri'euir'Q:~ . Jbe DT reaction requires the least amount of energy to initiate and is ~es:I as a s~lirce of IUglfenergy neutrons for boosting fission weapons and for thermonuclear~~RQ~~,}'
-;~~;t~~~::;;
~~=
.
...... '!!'
## E
i(::::~::;;:=-
.~~
(U) Electromagneti~ (~Ciio~~~~P~riiiion (EMIS) - A process that uses strong magnetic fields to separate ioiliz'ed uraniUiil~isotopes. The EMIS process works like a mass spectrof!l~u;t,J~Mt d~-~~~yf.h l~ger scale. EMIS was the major process used by the US
during:!Jie Ma.tlWittan P;oj.~t'for the production of 235U for the bomb dropped on Hiroshima.
lndiv1(J(ial uni~~were knoWn as calutrons, and some are still in service to supply small quantiti~(Qtyarious isotopes for research purposes. In 1991, the Iraqis were discovered to be developing')?MIS t~hnology for uranium enrichment.
-~~:;~~~~;~~
(U) Electron *,lA.* negatively charged particle which has a mass of about l/2000 of a neutron.
One or more electrons surround the nucleus of an atom. A positively charged electron is known as a positron.
(U) Electron Microscopy -A technique that uses a high-energy, finely focused beam of electrons to image samples at very high magnification. Features as small as 10 nanometers (I
x I0"8meters) can-be resolved. Two common versions are scanning electron microscopy
(SEM), which images the sample surface by reflection of the electron beam, and transmission electron microscopy (TEM), which images the interior of the sample by electrons that pass through the sample. SEM and TEM are analogous to reflected and transmitted optical microscopy. Many SEM and TEM instruments are equipped with x-ray detectors, which are
## Se~Forn
used to measure the elemental compositions of the samples. See Electron Microprobe Analysis.
(U) Electron Microprobe Analysis (EMPA)- A technique that uses a high-energy, finely focused beam of electrons to induce x-ray fluorescence in samples at very high magnification. The induced x-rays have energies that are specific to the elements in the sample. The measurement and analysis of the induced x-rays provides information on the elemental composition of the sample. EMP A instruments work like scanning electron microscopes (SEM) and can resolve elements at resolutions down to about l micrometer.
EMPA can detect elements at concentrations above 0.01 weight percent, depending on the element and the composition of the sample.
(U) Electron Volt- Also called an eV. A unit of energy often used in the m~~!#,~ment of n1,1clear and atomic radiation. One eV equals 1.60 x 10-19 Joules or 3.83 x to1'c~!~~es.~
. -~
.. ~z;:~;; .... , .. ~;
235U,Js~td~;~as b~~fi
(U) Enriched Uranium- Uranium in which the abundance of the increased above the natural amount (0.72 atom%). Most ligbt-water:~actors \l&~:~rimium enriched to 2 to 5% of 235U.
::;:
,..,::~:;;~..
?'
(U) Enrichment - The process of i~creasing the concentrati~il~~~;q~ i~:~~~:f an element
,relative to the other isotopes. In a typical enrichment P~gfe:s~, the ii~~a:i feed material is separated into a product stream (enriched) an<t~~~s (deP,J~red) stream for waste .
... ;~~,!~~~:~~~:=~~ ... '
'l:::::;-:.
(U) Event Class - A categorization of t!J..~~;~e of ~~~lear o~;~ai~logical event. The classes are: nuclear yield event, failed yield everit;:~:X.P.l~sive RDD event, non-explosive RDD event, source emplacement event, interdiction eveiit~:~;;.,
(U) Exemplar- A sample that serves as a.m~:i=~~;~~~dard. For nuclear forensic purposes, an exemplar would be a material of knowif'origin' and pedigree .
........
~ .. .:.....
.
.~ ....
. -~~ : ..
.. ;:~~;.'
.
..
:;~ ,.; ,.
,.<~-.
- {~:;
::i[i~:::;;::::(::
## F
----::-:.
.. /_.:.'.-.;~::~:~~: ./ti:_::--~-==-.: .. -
. .
.
.
(U)-F~~~t -.~itoome nJi~active particles from the cloud produced by the explosion of a
nuclearW~~p.im.
(U) Fissile;~=t::is~tope that can have its nucleus split, releasing a vast amount of energy.
~,. . .
(U) Fissile Material, Fissile Isotope - An isotope that readily fissions after absorbing a
neutron of any energy, either fast or slow. Fissile materials are 235U, 233U, 239Pu, and 241Pu.
235U is the only naturally occurring fissile isotope.
24~, and 252Cf (see
(U) Fission - The splitting of the nucleus of a heavy atom into two lighter nuclei. It is
accompanied by the release of neutrons, x-rays, gamma rays, and the kinetic energy of the
fission products. It is usually triggered when the nucleus is hit with a neutron, but in some
cases can be induced by protons and other particles or gamma rays. Some isotopes decay
spontaneously by fission, where an isotope naturally decays by fission without a neutron
trigger. Spontaneous fission can occur in isotopes such as 238U,
## Se~
spontaneous fission).
(U) Fission Products - The radioactive and stable isotopes produced by fission. Each isotope produces a unique pattern of isotope masses. In addition, the pattern of the isotope masses depends on the energy of the neutron spectrum. Three neutron energy ranges are of interest: fission spectrum (neutrons produced directly from fission), thermal (neutrons that have been slowed or moderated have thermal energies), and 14 MeV (which are produced by the fusion reaction of tritium and deuterium). Generally, the analysis of fission products focuses on the analysis of radioactive isotopes.
(U) Fission Spectrum Neutron - The energy of neutrons produced by the fission process.
Typical energies of fission spectrum neutrons are in the l to 3 MeV range.
,~i-
..
:::::: ..
(U) Fissionable Material - Commonly used as a synonym for fissile materiaft~;.rpeaning of
~-.
.
this term has been extended to include material that can be fissioned. by f~~,peutro~~Qp}y, such as 238U
;:;::;::;
::
.... :
.:i:~=r
:::::;f::/
(U) Fission Track Analysis - A method for finding particles that carifiin;uranium or
plutonium by fissioning a small amount of the fissile isotope ari'tt~.Qetectiit~;ftle fission tracks
''"-~"';-
--il"'
produced in a detector. Also called Lexan screening, beca~'Lex~
. J?lastic is used to detect
the fission tracks. This method uses optical microscopy,Jo"locate aiii;jsdlate particles of
interest in nuclear forensic samples. The pr~:'}~~-9~ ~,f!ij~~ipn tracks;:r~quires neutron
irradiation, which is usually done in a resear.Ch'te~ior.
.,*~;:.. . ..
.. ;:~t.
-~r
:;~;:::
(U) Fission Weapon- A weapon designea;w~pr9'fuce blast, thermal radiation, and nuclear
radiation through the fissioning of fissile miltilfi{l\1. (e.g., 235U and
23~). The complete fission
of one pound of fissionable material has a yield;~~y.alent to 8,000 tons of TNT .
.... ~.:Jr.~.
(U) Forensic Science - The comprehe~~i~{~~ieni~~c analysis of physical and biological
evidence in the context of civ,ih:~.J;.tliriill: or international law. The goal of forensics is to link
people, places,-things, and t:-.Yen~::see.also conventional forensics.
=~~ .~;fr'
.~-:~
(U) Fourier Transf'!r~1J:nt~~!"~~~(l#hR) Spectrometry -A high-resolution infrared
spectrometrY. !~cfmA~. that eniploys an optical interferometer to simultaneously detect and
analyze ~-wt~~f!n:ge.,.C>tiipfrared wavelengths. The data are analyzed by performing a Fourier
T~f9.hn, wii(~'h is a ~eiliatical process to convert intensity as a function of the
int~ci;~w.~te!~position to'mtensity as a function of energy (i.e., inverse wavelength). FTIR
becamef~~~tile with the advent of modem computers and computational algorithms. FTIR
is useful fd~~~.11JY3ing a variety of materials in solid, liquid, and gaseous form.
==~~r;-
(U) Fuel Eleni~~t - A rod, tube, plate, or other mechanical shape or form into which nuclear
fuel is fabricated for use in a nuclear reactor.
(U) Fuel Fabrication Plant - A facility where the nuclear material, such as enriched or
natural uranium, is fabricated into a ceramic material called uranium dioxide in a form
suitable for use as fuel in a nuclear reactor.
(U) Fuel-~ade Plutonium - Plutonium produced in nuclear reactors that has between 7%
and 19%
Pu relative to other isotopes of plutonium.
(U) Fuel Pellets - Typically, sintered (or fused) and ground cylinders of uranium dioxide, about 112 inch long and of various diameters, are stacked in tubes to form the fuel pins or rod.
(U) Fusion- The opposite of fission, in which two light nuclei atoms- typically deuterium and/or tritium- combine to form a heavier nucleus with the release of a substantial amount of energy. Extremely high temperatures, resulting in highly energetic, fast-moving nuclei, are required to initiate fusion reactions. '
(b) (3) per DOE
SRD, TCG-NAS-2
Topic 506.4
(U) Gamma Ray Emitter- A material that emits garrupji.:f.a~:(:~:~~!ler; photons). 60Co .
is a strong gamma ray emitter.
. ,,_.;:..
_X;~::::-.
_..
(U) Gamma Ray Radiation - Also call~9i:~=lli~tor s~~jtHi~-energy electromagnetic radiation emitted by nuclei during nucl~af:~~'?ti_!>rlS or radioactive decay. Many radioactive isotopes emit gamma rays and these gammatitY:~,have specific energies that are characteristic to the specific isotope and decay scheme. Gatillif~~t:_ays have high energy (generally 30 KeV
to many MeV) and a short wave length. Sltjelding;iigainst gamma radiation requires thick layers of dense material, such as lead. Q.afuma rays are potentially lethal to humans, depending on the intensity of.9leT~UX. :
.. -
_f.
-.
(U) Gamma Spectrom.':_t.ri~ . .A.detec,t}on system that measures the energy of gamma rays emitted by a radioac!i~~.>samp~~:=:Th~igamma ray energy and intensity is used to identify the radioactive isotope .and. the amount of the isotope in the sample. The detector types include solid-sta~:;d~t~tptS: "glis~P.roportional detectors, and solid scintillation detectors. The most COtnn:19P'type bf~solid-sta~~d'~tector uses a large, high purity, single crystal of germanium
(also=~liJ!~~ IWGe). A soii'il scintillation detector often uses large crystals of sodium iodide.
The mosf~~~itive gamma spectrometer can detect activities of 0.1 dpm or less.
!";:;!::,...
.
(U) Gas Ce~t6fuge - A uranium enrichment process that uses rapidly rotating cylinders (gas centrifuges, also called rotors) to enrich 23Su. Uranium hexafluoride (UF6) gas is fed into a gas centrifuge and rotated at high speed. The centrifugal forces produced by the rotation cause the heavier 238U to migrate toward the outside of the rotor and the lighter 235U to migrate towards the center. Gas removed from the center is slightly enriched in 235U and gas removed from near the outside of the rotor is slightly depleted in 235U. The separation efficiency is increased by a relatively slow axial countercurrent flow of gas within the centrifuge. The countercurrent flow process produces further enrichment by concentrating enriched gas at one end and depleted gas at the other. Feed UF6 is introduced near the middle of the rotor, and enriched and depleted uF' are removed near the ends. The separation capacity of a single rotor increases with the length of the rotorand the rotor wall speed (i.e.
with increasiD.g rotation rate). Consequently, centrifuges consisting of long, high-speed
## Secre~Rn
rotors are the goal of centrifuge development programs, subject to materials and mechanical constraints. A typical gas centrifuge plant can have thousands of rotors, connected in series.
Gas centrifuges must be constructed of very strong materials, such as carbon fiber or highstrength aluminum or steel alloys. The gas centrifuge process is much more energy efficient than gaseous diffusion.
(U) Gaseous Diffusion - This isotope separation process is based on the fact that the lighter isotopes of 23SU gas diffuse through a porous barrier at a faster rate than the heavier isotopes.
This method requires large plants and enormous amounts of electrical power. China, France, Russia, Great Britain, and the United States have used this isotope separation process.
(U) Gas Proportional Detector - A radiation detector that detects beta particles o.r gamma rays by the ionization they produce in a gas. Gas proportional detectors can ~~~~re the energy of the gamma ray and are used in gamma spectrometry. The most sen8itiY.~;.gas .
proportion detectors can detect activities ofO.I dpm or less.
,::;.
-:::;~;;.;,;;
(U) Geiger Counter - A radiation detector that can detect beta partic1et:~~~~fnma ;~;s.
The Geiger counter is widely used for radiation detection trl~asureni~if~~(qr he~ifu. physics and radiation safety.
;;!:::.
'::!;~:;~<
.. :
.. :..:~~.. .
: .. ,.-
(U) Glow Discharge Mass Spectrometry (GD-MS) - :fu ~i~w d~~~lifg~ ~s spectrometry
(GD-MS), the sample serves as the cathode o~~,gJ.ow ~~~~!lfge (argon is usually the support gas). The sample i~ sputtered by argon io:J?,~;iftiCi::ij~.spurti~r~~t~~.utrals from the sample diffuse into the plasma. In the plasma, tlt~t.:~eutral~{~ ionize~l'ieither by electron impact or, more typically, by collision with metaSt3.Bi~:M:g~n:=iitoms (penning ionization). GD-MS can be an effective technique for directly measuii'Q~tbulk samples, such as dirt. GD-MS is highly quantitative, suffering from very few matrix eff~~~-.-Jtcan be fused as a sensitive survey tool with detection limits ranging from less th~;11 ppiY:i&'a few ppm, depending on the element.
However, it lacks the precision associat~5f!\Vith railiochemistry, TIMS, or ICP-MS. It also can provide misleading result.~.~f~~;~om~'heterogeneous samples, since the sampled volume is small, and there is no sampl~'hopJoge:Q.ization provided by dissolution or a similar process .
. .~.
==~ ... ~--~
.~::
(U) Graphite - A forni:o'f catb,o.~.;:Effiiphite is used as a neutron moderator in some nuclear reactors. Such react~~. can ~-()ifii'atural uranium and are useful for producing weaponsgrade pl~~ffi~} M~~~t-~he_ US weapons-grade plutonium was produced in graphitemode~~ed reac(Qrs at th~~WSDepartment of Energy's Hanford Site in Washington state.
_
_..:::;:~:i:;,,
. ::--
y
(U) Greeg:~~t- Green Salt is the term used to describe uranium tetrafluoride (UF4) which is a green ccy~f@!:W~.solid compound of uranium. UF4 is generally an intermediate in the conversion of\lilihium hexafluoride (UF6) to either uranium oxides (U30s or U02) or uranium metal~' It is formed by the reaction ofUF6 with hydrogen gas in a vertical tube-type reactor or by the action of hydrogen fluoride (HF) on uranium dioxide.
## H
(U) Half-life - The amount of time needed for half of the atoms of a radioactive material to disintegrate or decay.
(U) Health Physics - The science concerned with recognitioD, evaluation, and control of health hazards resulting from ionized radiation.
(U) Heavy Water - A form of water in which the hydrogen has been replaced by deuterium.
Heavy water is used as a neutron moderator in some nuclear reactors. Such reactors can run on natural uranium and are useful for producing weapons-grade plutonium, Some of the US
weapons-grade plutonium (and most of the tritium) was produced in heavy-water-moderated reactors at the US Department of Energy's Savannah River Site in South Carolina.
(U) Highly Enriched Uranium (HEU) - Uranium that contains 20 _atom% or more 23SU.
(U) High Explosives (HE) - Energetic materials that consist of chemical com~Qtiiids or mixtures of compounds that when properly initiated evolve large volumes of ilfs:IA:!i short period of time and produce_ a high pressure shock wave that is capable o(~~!J:erixl8;P.r:.;;;
compressing surrounding media. Examples of HE compounds suitable:fdt'ii'ti~I~ar w~apons include:
:~::
/~~i:..
-;~:!:1~;::::
(U) HMX (cyclotetramethylenetetranitroramine)
:~:::..
-::::~;;;_:,.<
~
~~(~~~':a~:!;~~~:::::;nne)
,:;:;./:: . :;:~~~~~~~)-... /-
(U) TA TB (triamino-trinitrobenzene)
.;.::..
_:-'-::;~:_;_::....
"
(U) TNT (trinitrotoluene)
. /;,:::::.~:.;:_:
(U) Tetrya (trinitrophenylmethylnitr~w~)
... ,.
-\!::=-
.. il..
..:.
._,.. ' .. ~~::;:~:~ .. -~-
(U) All of these HE compounds consist of org@)~ compounds with attached nitrogen oxide
(or "nitro" groups), which are responsible for tfie;~~.P.~osive character of these compounds.
For nuclear weapon applications, these (or:~imilaii:C!ompounds are usually blended with inert binders (such as plastic) to obtain a phY.~!.tkl fom1ofthe HE compound that can be molded or machined into the desired shaP,:;:;:::.
::
... -
. .. ~~
.
;i
_'!
l
(U) HPGe Detector - A,.~,.ef g~
ray detector that uses a large crystal of high purity germanium to detect g!ilffina'~Y,~.:":HPGe detectors can measure the energy of the gamma ray and are used.i.t! g~,spectroliie-fiY. HPGe detectors must be cooled to cryogenic temperawrC:S;~~p~h:'as'li~;~ing liquid nitrogen, to work properly. Hence, HPGe detectors are not e~jiy mad~!f;ortabll~~:}Jte most sensitive HPGe detectors can detect activities of 0.1 dpm, or les~:~;;:. . }
_,:
-.-;" .. :. ..
:::::~=:
-.. ~i!.'a
.. :;~:;:::~.:-; --
~:::
/
## 1
(U) IC- Intelligence Community.
(U) ICP/MS- See Inductively-coupled-plasma mass spectrometry.
(U) ICP/OES - See Inductively-coupled-plasma optical emission spectrometry.
(U) International Atomic Energy Agency (IAEA) - An independent, intergovernmental, science and technology-based organization of the United Nations that serves as the global focal point for nuclear cooperation. The mission of the IAEA is to verify through its inspection system that its member States comply with their commitments under the Non-
Proliferation Treaty and other non-proliferation agreements, to use nuclear material an~
facilities only for peaceful purposes.
;:::"'
-:;;;:.~ .. -- .. ~
.. ~
"' .. ".'"~~>'""
(U) Inadvertent Signature - A signature, su~h,a,s .. ~ tr11ritii~J~ment or.i~otope that is present in a material that is othe~ise unimpo~t to .~~=ti~e;~~~ the ~~~B~ An example would be ~
trace amount ofplutomum found wtth an':HEU sample. This"trace Pu would have no bearmg on any of the material properties ofthe'Irniit;I!9t.does its conbentration in the HEU pose any issues for its bulk nuclear properties. The tili~~eP.,u would not be controlled, but it would be a signature of the process that produced the HEif;~:~:;.>/
6 ........
.
. .
. -~:.~:_?:-
(U) Inductively-coupled-plasma Opti~aP:iJtmission Spectrometry (ICP/OES) - An instrument used for elemental.~ysis~tfiat uses a hot plasma to vaporize elements for optical spectrographic analysis. ~{nq~almethod of introducing the sample into an ICP/OES is by dissolving the sampl~! g~~futing.~ aerosol from the solution and feeding the aerosol into the hot plasma. The h~Fpla8ili~:sil#~s many of the elements to emit light at characteristic wavelengths. The s,p~p-ograplii'c'analysis of the emitted light is used to determine what elelllen~,a.re:~t,e.~,eiit Iii:i~s: sample and their concentrations in the sample. ICP/OES uses the
.:.
~-- .. ~
~!i :II;..
......
..
same,m;e ofpl~ina torc~Ji~~l.lsed man ICP/MS. ICP/OES can detect elements down to partper.-b'iilion to pan-per-million levels.
-~~:;:=~~;, *'1::* J
-;-~~
(U) Inducti~~h~:-eo.upled-plasma Mass Spectrometry (ICP/MS)- A mass spectrometry technique th~tf@~~ a hot plasma to vaporize and ionize elements for mass analysis. The normal method"ofintroducing the sample into an ICP/MS is by dissolving the sample, generating an aerosol from the solution, and feeding the aerosol into the hot plasma. ICP/MS
was initially developed in the early 1980s and has been developed into a general technique for analyzing the elemental compositions of samples, as well as the isotopic composition of individual elements. ICP/MS is pmnarily used for bulk analysis. The technique normally requires extensive cheinistry to dissolve the sample and to separate and purify the element for analysis. However, for particle analysis, small, micrometer-sized particles can be loaded dir~ctly onto a filament for isotopic analysis of uranium and plutonium- no chemistry is used and the atoms are vaporized directly into the carrier gas and transported to the plasma.
ICP/MS.can detect as few as 100,000 atoms ofplutonium, and can measure isotope ratios to better than 1 part in 1000 precision for larger samples.
(U) Ion Microprobe Mass Spectrometry - A mass spectrometry technique that uses a focused ion beam to remove (or sputter) atoms from a sample for analysis in a mass spectrometer. The ion microprobe can analyze the elemental and isotopic composition of
small pieces of material without the need to perform any chemistry on the samples, which is different from other types of mass spectrometry. The technique can analyze areas smaller than 1 micrometer (or micron) in diameter, can detect elements at concentrations as low as one part-per-billion, and can measure the isotopic compositions of elements to precisions better than one part in I ,000. Also called secondary ionization mass spectrometry or SIMS.
(U) Ionization - The removal or addition of an electron from an electrically neutral atom or molecule, thus leaving a positively or negatively charged ion, respectively. , :;:: -- .-~
(U) Ionizing Radiation- Any radiation that causes the removal of electrons :ff~iq~~~oms.or molecules, thereby producing ions.
. .....
~;;:; .. ,,::
;::~~:::!~,~.
. .;:;:,.
(U) Isotope - Atoms of the same chemicai element but wi~.;Piffereq.ti~~b;~;~ltJ.eutrons in their nucleus. An isotope is specified by its atomic weight~aiid a syin'I)Qlidenotirig the chemical element, such as 235U for uranium with 235 neutrons.ahd prot~~~iJs"Otopes can be either stable or unstable (radioactive).
,.;,, ;.-;.::;:\.
}i'
## K
(U) KeV- Kilo-electrQii:~ott~:l;Q:ikctron volts): a unit of energy often used in the measuremen~ p_f nu~J~~ radiatfoii=such as gamma rays, alpha particles, beta particles, and neutron!!, .. anO::~~Qnttc.hlaiation such as x-rays. See also electron volt and MeV .
.
<~=;i.'
~~=~~:
-.~~:~;;:;~ .. ;
(U)l(~q~n S_~fuple (K) :A sample of known origin and attributes, which is used in the forensic~C:<iqiparison with an unknown or questioned sarilple. The known sample has been previously~abalyzed and documented and the information is likely to have been incorporated into a datab~~~~~See Questioned Sample (Q).
,.
(U) Krypton (Kr) - A chemical element with atomic number 36. It is a noble gas that occurs in trace amounts in Earth's atmosphere. There are 20 known isotopes ofKr. Naturally occurring Kr is made of 5 stable isotopes and one slightly radioactive isotope. 85Kr is a radioactive isotope produced by the fission of uranium and plutonium and is produced by nuclear reactors and nuclear explosions.
## L
(U) Lanthanide Element -A group of elements (also called the rare-earth elements) on the periodic table of the elements that includes lanthanum, neodymium, cerium, and samarium.
These elements are chemically similar and have been grouped with the lightest element of the series, lanthanum (hence the name lanthanides). The lanthanides are important to nuclear forensics because they include many important fission products, which can be interpreted to determine the fissile isotope and to determine the energy spectrum of the neutrons that produced the fission. The lanthanides are also similar to another group of elements, the actinide elements.
(U) LEA - Law Enforcement Agencies.
~=:~ ....
.. ~
1;.1'1;." ..
(U) Liquid Scintillation Detector - A radiation detector that detects alpha pattie"\e~ or ~ta particles by the light they produce in a liquid. Liquid scintillation detec~f;.~an mg~W:~~the energy of the alpha particles or beta particles and are used in alpha speji6ih~~. The'inost sensitive liquid scintillation detectors can detect activities ofO.l dpni;o'r less. {::~, - ..-
,:~:>; '
;::;~=~~::.
}''
(U) Lithium - A chemical element that has nuclear properties .$~tare usethl,for thermonuclear or fusion reactions. Lithium is also used ~.,3;-~rgei';(~,r production of tritium .
(U) Maraging Steel - Maraging steel is an:4mpo~r~omponent in the design of gas centrifuge rotors. It allows for the veryJpgh rotor" wall speed necessary to separate 238U from
235U, This type of steel (whic!l:b:~::a htgb cobalt content) is the most popular rotor material for proliferant countries to lJ.$e it);fiuilqing isotope separation facilities .
. - !;: ~~/~
):
(U) Mass Spectrom~t~f;l Ah~!!q~~~l instrument used to measure the composition of a sample based on the~atomic (or'blolecular) weights (masses) of its constituents. Some mass spectrorp,.~ilit$~~~ h~~t~ly measure the isotopic composition of samples .. Different types of mas&spectrometers includf:
.... ,,.,.
~Iii!'
;~"'
.
(U)-~~~~lerato.r'mass spectrometer- AMS
(U) Induct!:V.~!y-coupled-plasma mass spectrometer - ICP/MS
(U) Ion mi~9-P.~pb~ mass spectrometer (also called a secondary ionization mass spectrometer
- SIMS)
. ~:::::;,
(U) Thermal irlriization mass spectrometer - TIMS
(U) Each type of mass spectrometer has a specific range of elements (hence isotopes) it can
analyze and associated detection limits, as well as requirements for sample preparation
chemistry and handling.
'
(U) Material Signature - .. Material signatures" include all characteristics of a particular
material, whether the characteristics ate specifications or nor - i.e., the material
.. fingerprint." .. Material sigriatures"' include ''process signatures", but also include
inadvertent or unspecified signatures. Inadvertent signatures are unimportant as
specifications for the process and, hence, are not under deliberate control. Evaluatidn of
material signatures has the potential to identify where (hence, by whom) the material was made. An important point is that "the interpretation of material signatures requires extensive databases to reveal the origin of the materials." "Exemplars- that is, examples from known processes and locations- are essential." See Process Signatures.
(U) MeV- Mega-electron-volt (1,000,000 electron volts): a unit of energy often used in the measurement of nuclear radiation such as gamma rays, alpha particles, and neutrons. See also eV and KeV.
{U) Micron - A micrometer or one-millionth of a meter ( 1 x 10-6 meter). See micrometer.
_(U) Micrometer- One-millionth of a meter (1 x 10-6 meter). The wavelen~.ofY.ellow light is 0.5 micrometers.
i;::/'
.. .~~~ ,..
{U) Molecular Laser Isotope Separation (MLIS) - There are two basic .. ~~~~ ~~~~l~9;1n the MLIS process. In the first step, UF6 is irradiated by an infrared las~fS'y~i~ ope~iing near the 16 mm wavelength, which selectively excites the
2~-:UF6 , lea~g the;~~6
relatively unexcited. In the second step, photons from a se~6nd 1as<;a~~~m (infrared or ultraviolet) preferentially dissociate the excited 235UF6 to form}?~UFs aifa~it;ee''fluorine atoms. The 235UF5 formed from the dissociation precipita~N:.~.::frotii:~m~ gas-iiS' a powder that can be filtered from the gas stream. In terms of the gas ,~qw for the~~IS process, gaseous
235UF 6 is mixed with a carrier gas and a scavet:tg~F .. gas Js:~~panded th'fough a supersonic nozzle that cools the gas to low temperatw;~s:~;Hyaf.Qgen Br;:i,ooble gas are suitable as carriers. A scavenger gas (such as me~.~~~ is us~io captt.if~:tb.e fluorine atoms that are released as a result of the dissociation of
23~tJF.6 niolecules. Like AVLIS, MLIS technology appears promising bus has proven to be extr'~fEiy difficult to master and may be beyond the reach of even technically advanced states.
~:::~::~:-:/
.
;~: ...
(U) MOX-Grade Plutonium - Mixed Q.ilc:ie foe( which contains both uranium and plutonium oxides. MOX fueljs::m.~e ~ing plutonium extracted by reprocessing spent power reactor fuel and typically h~'~ier t'tt_an 30% 24<Tu.
!~~ -~~'!
~!:
foa ~-
"lo"
-~~~;=:~ =~~;::~::~~.
Atoms are about
(U) Neptunium (Np)- A metallic radioactive element with atomic number 93. Neptunium is found in trace quantities in uranium ores and is also produced synthetically in nuclear reactions.
(U) Neutron - The neutron is an electrically neutral particle of nearly the same mass as the proton. Neutrons are one of the three basic particles that make up the atom, the others being protons and electrons. Neutrons as released by fission and can be produced by other nuclear reactions, such as bombardment of beryllium by alpha particles. Neutrons can be absorbed by paraffin, hydrogenous material, or by very thick layers of lead.
(U) Neutron Initiator - A device that emits a burst of neutrons to start a chain reaction in a critical mass of fissile material. The initiator triggers the nuclear detonation once the critical mass has been assembled by high explosives.
(U) Neutron Moderator - A material that slows neutrons. Examples include graphite and heavy water. Moderators are used in nuclear reactors to slow neutrons from fission energies .
to thermal energies to increase the probability that the neutrons will react with the nuclear fuel in the reactor.
(U) Nickel (Ni) - A chemical element that is used in many nuclear applications because of its chemical resistance to oxygen and fluorine. Nickel is also a component of many types of steels.
~~:..
~ <~
.. ~ .. '~'";.. .
(U) Nuclear Attribution- The assignment of responsibility (people) for the fii=i~~~~d or.
actual use of nuclear material.
.:~>.
':~!;1:;.,)
:~=~'- --~:~~~~~..
~: .. :~::"
(U) Nuclear Characterization- The description of the ch~tp.ical, el~]ri~ntal,'
::~pic, and physical aspects of the nuclear material as well as the inferf~ histoilaq:~(pedigr(!eJprocess)
origins of the nuclear material.
.:=::::..
:::;:~:.~~ ..
. .. . .
~
~:~:~:::.,
?t~
(U) Nuclear Forensics- The application of nuclear physici~:~nd ~~iy~cal technologies to the study of nuclear material in forensic samples .. M:~ith;:6~qyentionaLforensics, the goal is to link people, places, things, and events. Nu_~l~::r6~sics it~~.s~ks to identify how nuclear materials were produced, their intended u&e, and W.here they"we're produced. The findings may be presented as technical evidencliiri~~~~r:Sf law or oh. a national security setting.
-~ .... - ~li
(U) Nuclear Detonation -A nuclear explosion:~~~!~ng from fission or fusion reactions in nuclear materials, such as from a nuclear w~apon:;:
.-:-;....
.
(U) Nuclear Energy - The energy::r.etei~d when the nucleus of an atom splits or when two nuclei fuse. (See fission an~=fu~ion). : .
.
~..
:~;~;::~
'~~:
(U) Nuclear Radiati_o~:~'Pai'ij~~~;~d'electromagnetic radiation emitted from various nuclear processes in .~t~mic;;P~~lei. The'imiJortant radiations, from the nuclear weapon effects standpoi9t;i~~~lplia 'liii<J::beta particles, gamma rays, and neutrons.
(U) Optical Microscopy- A microscope that magnifies light. Features as small as about I
micrometer (or micron) can be resolved by this technique. Two variants are reflected light microscopy, which images the sample surface by reflection of light shone on the sample, and transmitted light microscopy, which images the interior of the sample by light that passes through the sample.
(U) Parent Isotope - A radioactive isotope that decays to produce a daughter isotope. The concept of parent isotope and daughter isotope are crucial for age-dating.
(U) Penetrating Radiation - External radiations of such penetrating power that the absorbed dose from exposure is delivered in significant and damaging quantities to human tissue and other organs. It refers to most gamma radiation, x-ray radiation (excluding those with very low energy), and neutron radiation assembly, but not including safmg material.
,
"~....
~
:!'.,'rl~ ~'
(U) Pit - A pit is the core of an implosion type nuclear weapon. It contains th~:;Kssile material and any reflector or tamper associated with it.
:~~;:::,~ ... ::
;;f ~;~:~ :=:~...
~::.~ ~ ..
lliJ Plutonium (Pu) - A transuranic element with a fissile j_~.9tope o.~~~ss.n~~~f239
(23 Pu).
~-:-' .::::~. ,,. -;~~;:~:::~:~:/ /
(U) Poloniu~-210 e'0Po)- ~alpha-emitting radioactiv~~~~blp~:~}h ~ J~g day half-life ..
2'0po occurs m trace amounts m nature due to the decay.ofnaturally;oooumng
23~. 2'0po ts produced in nuclear reactors by irradiation of
2~:i (th~s9J~.stable is<Stope of bismuth) and is used commercially for a variety of purposes,t(sii6Ii:ijs.. in srii~:fi;,_encapsulated sources to eliminate static electricitY) and in nucle~~eapon~:;iiS part of~;ii~utron initiator (when mixed with beryllium).
./ ::::;:;....
/~
'
4::~~:=~=:
(U) ppb - Parts per billion; refers to the con~~~~p~of something in a material.
.
-~:.t.~
.~;j~-
.:
(U) ppm - Parts per million; refers to ~e::ooncenti"ation of something in a material.
.... ~ ..
~ ~ .. ;.,..
~.-.. ~
.
(U) ppt - Parts per trillion; re:f~;ji~-~h~ concentration of something in a material.
.
~: ... !;:,
.. ::
(U) Process Signatur4!"r~? P1~b{~i~~tures generally arise from product specifications,
which depen~ ~n h9w~:iJte matefiiii'was produced and how the material will be used. An importantpil'iJii:i~ that:~t};rpcess signatures can generally be interpreted without the need for exte~~~ye data~~es." rri~wtelpret process signatures one usually needs only knowledge of the.pf~,s (L~~'fmal prod'bct specifications). See Material Signatures.
..
.. -.. -.......
~~
.
(U) Prot~~:~;~~ .P.roton is a positively charged particle of nearly the same mass as the neutron. ProiB:~;Iire one of the three basic particles that make up the atom, the others being neutrons and eiectrons.
(U) Pusher - A shell made out of low density metal- such as aluminum, beryllium, or other metals- which is located between the explosive lens and the tamper in an implosion type nuclear weapon. It works by reflecting some of the shock wave backwards, having the effect of lengthening its duration.
## R
{fooQ) Radiological Dispersal Device (RDD) -A Radiological Dispersal Device (ROD) is defined by the US Department of Defense as, "any device, including any weapon or equipment, excluding a yield-producing nuclear device, designed to employ radioactive material to cause destruction, damage, or injury by means of radiation produced by decay of such materials." An RDD may cause mass disruption- possibly causing mas$. hy~Jeria, fear, and significant cleanup costs- but is not considered a WMD
:;:;,(
~=~~==!i:::..
-
(U) Radioactive Counting Techniques - Each radioactive isotope emi~SM\jatiorie~~~own types and energies at a known rate. By measuring the radiation emitte(,lilifhts.~ple;ft'is possible to quantify the amount of each measured isotope P!-~sent. ~~~e are ~~-types of radiation that are usually considered for measurement.: alplia~ beta ahog~ma nRtiation.
Each type of radiation has its own properties and methods of ru;ieP.tion. ';;:;;::
(U) Silicon surface barrier detectors commonly detect alp~~?pidiacll~!!:
}"'
(U) Scintillation techniques or gas ionization detectors ~~(used to d~fbeta radiation.
(U) Germanium crystals are commonly used ~q,9:~~,ct .8'1iWwa radiation.
'"~::!;(~:::~::::;: -
.-::~~~=~~~~ .r'
(U) Radioactivity - The phenomenon, ~~~ited \>Y.!and bems::aproperty of certain elements, of spontaneously emitting radiation resultiqg.from':Changes iri the nuclei of atoms of the
.... ~: .. :~
element.
..;.
=~=~=~-
.
.. :::::;::.~4
(U) Radiochemistry - Many samples are ,t~,9 com;lex for all the radioactive isotopes present to be measured directly. By utilizing fl!~;idifferenees in chemical properties of the elements, it is. possible to devise schemes.9-:~!,l:emi~iil reactions to separate and purify elements, or groups of elements, to allow measurem~ni oft}le isotopes present by radioactive counting methods, or mass spectrometry. The fs.otf>pes l!(easured are related back to the original sample by referencing to an interriai'isrlf<>pjc;.sWi'dard called a "spike." The chemical separation and purification steps intYcltse botii'tlifsensitivity and selectivity of the technique.
Radioch~~~~ftY;;is ~~~~Wly important to allow measurement of isotopes that are present at low a~P.t'ity an'&,r!ire besi~~ured by their alpha or beta emissions or by mass spectrometry.
~i\~i~::
/='
Raditi~li~~istt:Y'J.n combination with radioactive counting techniques and.mass spectrometry has the.fj(;wp~al to measure down to 106 atoms or lower of certain isotopes.
s
(U) Sating Arming Fuzing Firing (SAFF) system - A SAFF system is crucial to the development of a militarily usable nuclear weapon. It consists of the following subsystems:
(U) Sating: Used to ensure that a nuclear weapon will not experience a nuclear detonation as it is being stored, handled, deployed, and employed. Sating usually involved multiple mechanical interruptions of both power sources and explosive firing trains. The nuclear components may be designed so that an accidental detonation of the high explosives is
intrinsically unable to produce a significant nuclear yield.
(U) Arming: Placing the nuclear warhead in a ready operational state, such that it can be initiated under specified firing conditions. Arming generally involves mechanical restoration of the sating interrupts in response to conditions that are unique to the launch or deployment of the system.
(U) Fuzing: Detecting that the desired conditions for warhead detonation have been met and providing an appropriate command signal to the firing set to initiate nuclear detonation.
(U) Firing: Delivering a precise level of precisely timed electrical or pyrotechnic energy to one or more warhead detonating devices.
=~=:==~~,
,.
- -
(U) Tamper - A shell surrounding the fission core in an imnlosion tyP.i~::~~~~~pon which keeps the nuclear material confined during the impldsion for"a'Iij~ger timtraising the
. ld f th
. - .
=~~~- ...
yte o e weapon.
;.;.;-,
,=~-:
.;:~::,., =:t~~;..
?'
(U) Thermal Ionization Mass Spectrometry (TIMS) 7.1ii"tiMS, a::~~ple is deposited on a metal filament, w~ich _is hea~d i~ a high ~ac~~:~:,; ~~~~~-a currerifthro~p it. TIMS is capable ofmeasunng tsotoptcranos on ptsggram<(:!~9
12 g)tQ~~~c;>gram (10 g) samples, or down to tens offemtograms (1 x 10"15 g);,p~jng spe~jil pre-c6!)"c-entration techniques. TIMS
routinely measures differences in isotopc(riiiiss ratibs of the order of 1 in a million.
{U) Thorium (Th) - A radioactive metallic :~J*~~-.vy.ith atomic number 90. Thonum may
. be used as fuel for special types of nuclear:t.:eacto~~
.. ~!;~:;'
.
(U) TNF- Technical Nuclear.F.ort:nsics:
- I
~,~.~
..
.
....
: ... ./
::::~:=
\~i~::::::/:::-
'!.II':~"'
;:~-:
U
_
....
.. /("'lt:~
"~''\;--'
M. U;~~~ttf~U) -A ra~oactive element with various fissile isotopes. 235U is used in
nuclear witl~ns and as fuel for nuclear reactors.
-~::::~~-
(U) Uranium-ilexafluoride (UF6) - A compound used in the uranium enrichment process
which produces fuel for nuclear reactors or highly enriched uranium for nuclear weapons.
UF 6 is used as the feed material for gas centrifuges, gaseous diffusion, and MLIS enrichment
methods. UF6 is produced by reacting UF4 with fluorine.
v
(U) Visual Inspection and Photography - Visual inspection of a sample can give an expert
## S~Rn
information as to its possible identity, especially in conjunction with data from NDA
techniques such as gamma spectrometry and survey data. Size and shape can be sufficient to identify some items, especially if serial numbers or other identifying marks can be seen. For chemicals, the color and form of the material can be important clues.
(U) X-ray - A form of electromagnetic radiation, similar to.,isible IJ!Wl but of.~~w:ter wavelength (between 0.1 to 10 nm) and capable ofpenetraiftig solicfs~~.jonizipg gasses.
!"~~;;~~~t
~:::;j~:~~? .\
(U) X-ray Diffraction Analysis (XRD) - XRD analysis isJ!J.' s~~d method for identifying the chemical structure of inorganic and orgat)iS'ciystallin~:!;Iiaterial. X-ray beams that impinge on regularly ordered lattices underg9-,~on~~tive and d~structive interference that depends on the spacing of the lattice, t:Pe'Wii\i:fJ.~gth6f;J9F.?C-rays, and the angle of incidence of the X-ray beam. By rotatingji}.e samP.\e relative:tf{a fixed X-ray source, variations in interference occur, leading"t6~~~,~teristic ditrraction patterns. These diffraction patterns can be compared to refert?~~.~pectra to identify the specific crystalline phase. XRD cannot generate diffraction pattems:~~~~amorphous (non-crystalline) material.
(U) X-ray Fluorescence Analysis <XI9.9;~::XRF .~:lysis can be useful for the broad and non-destructive elemental qWlP.tifig.atioii of a sample. An incident X -ray beam is used to excite characteristic seconrupy ~-fay y.:avelengths and energies in a solid sample. These X-
rays are counted using ~.soll,q.state o:r:;proportional counter. The detection limits for XRF are in the range of 10 p~~~!i"i\mil*~i.~:8f~tlie light elements is possible but more problematic du~
to the low characteriSti~ X-ray energies involved. XRF is strictly an elemental analysis tool, while IC:p.,-M~:9.r. d:i:i~~S.. which are more sensitive, are able to measure isotopic compqsition. ')tlzy can fi=~:irefformed direetly on solid samples, although dissolutions are ofteri;;il'ii~!yz~:ii> providehomogenization of the sample.
~ .&
"' "'
(U) Xe~~~~&.~l- ~ chemical element of atomic number 54. Xe is a noble gas which occurs in the Earth's'il:iQiosphere in trace amounts. Naturally occurring Xe consists of9 stable isotopes. Xe afso has over 40 unstable isotopes that undergo radioactive decay. 135Xe is produced as a result of nuclear fission.
## Y
(U) Yellowcake- A processed oxide of uranium (U30s) which is extracted and concentrated from uranium ore. It is used as the raw material for commercial nuclear materials.
## Z (U) Appendix C: Dynamic Analysis Process
(U/tRmO)Very hard to nearly intractable intelligence problems usually go through to an initials~ by a highly-focused special analytical team. Outputs from the team's efforts are used to update collection requirements, and sometimes result in the formation of a new entity to work the problem full-time. After a long period of little progress or no real intelligence production, the problem may fall below the day-to-day focus. If and when activities related to the concern about the issue remind consumers of the threat, then another focused look at the problem may be initiated.
(U//f'6:t.JO) For certain classes of issues, failure represents catastrophic conseq1,1~nees. Issues of this n~e demand a level of focus that digs deep every day, even if the resiilts-:are negative over long periods of time. In such situations, dedicating some reso~~!:fiJ;.wodcing the problem on a continuing basis is worth the effort.
:;:i:;;::::...
~:;:;:::'
'
-J<."' .. ~
.
~
~~~---~:
~-~:!=~~~ --:~--
~U//F~O) A dyn~ic approach to analysis ~d collect~ori=t6r bard:{~:~~ly in.~ctable Issues ts Illustrated m the chart below (Dynamtc Analysts and ~-~!Jectton'~s:le Process).
This approach has previously been recommended in severa!;stUdie~!perforliied by the DNI's now defunct Intelligence Concepts Development Office..({cirmerly 'ii~~~r'the ADCI/Collection as the Collection Concepts Development Ce<n~c:9..-:-Jh~.;~~~endatibns were generated by applying the process to developing analysi~._aiid~dH~ctioti=~qQmmendations on particular hard problems identified by the NICB. ..:;:::
-~~::~
::;7;_;;.'
.
. .... -:~;:;z: .. _ .... ~~7
~
(U//F~O) The concept for dynamic anal;~i~1i:f~-~ig~t forward. The first phase involves creating a notional series of phases that any entity;;}Xoidd use to start and achieve a particular objective. The actual steps would be b~;:on wl)iifever is known about the ilk of the entity and real world activities that would have;:ib take place to achieve their objective .
. .. .;-....
~ ..
.. ~~;::--~:~;:; ..
(U//FDuO) Under each ph~~~;~i'step~1~articipants in the process brainstorm the manifesthions of activities ttiat .. must::occur in order for an entity forward on their objective.
In concert with this .l;J~tonrlirig;=p-iuticipants identify what might be collectable from any particul~.-9'!~t~stifi6~::.JFor example, there would be a number of areas where information and p}?.~itomeli*~ould -ti~~g~erated in the course of arranging for a safe-bouse to set up a cov~;ip~eting.:dtthe ''grriuP..wbo-hates-potatoes". Someone would have to make the aria~g~~~p.t:sfor a room or house-leaving a trail of paper. Some attendees might rent cars-anriii!~~.!fail of paper. Cars on the way to the meeting would cause a Doppler shift in any local radi~]~~quency signals. An upswing in anti-potato rhetoric might show up in internet medi~f Possible collectable information would be arrayed against the capabilities of existing collection resources.
(U//F~O) If a collection resource had the access to and ability to collect the information, an approprilh:e requirement would be initiated. Alternatively, requirements already in place would be updated if necessary.
(U//F~O) If collection was feasible but no capability yet existed, an associated gap-closing research objective might be initiated.
(U/F~) When conducted on a conti uing basis, the dynamic analysis process would SEC
OFORN
constantly iterate the issue's evolutionary steps and possible manifestations. These iterations would be based in any refinement in understanding the ilk of the target entity and specific intelligence information. (A simple analogy is the constant process of writing and rewriting scripts for a movie, as a movie director shoots and evaluates scenes or as the script writer does more research.)
(U/ ~)
Work on nuclear attribution is more than just picking up the pieces after an event takes place. Vigilance in working the issue before any harm takes place is a vital part of the job. Interdiction as early as possible should be the primary objective. Some resources should be dedicated to a achieve the objective of frequently revisiting what indications might emerge in LE, TNF and intelligence data from motivation to the aftermath of a terrorism-based nuclear event.
## S~ 0) Dyamic Analysis And Collection Cycle Process (U) Policy Documents
(U) END NOTES
:-. ..:;
::::::::
.':"lele.;::'.;,
. ., :~::~:~..
~
1 [Journal I Weiss, Charles !"Communicating Uncertainty in. Intelligence and Other Professions" I f~lt?inational Journal of Intelligence and Counterintelligence I Date of Publication: 2008 I Pages 57f8"5"j'V.~!~:.? II
. ., .
2 [Journal I Weiss, Charles I "Scientific Uncertainty in Advising and A~y.!,>Cacy" I J:~~nology'iij;}~pciety I Date of Publication: 20021 Pages 375-3861 Vol. 24]
.
:
~ ::~:~:..
_.:~:
3 [Website 1 Kloprogge, Penny; van der Sluijs, Jeroen; and Wardekker, ArjaA:t::unce.:Gqi~.Co~munication Issues and Good Practice"! December 2007 I
.... '<:1:::;..
:('
URL: www.nusap.net/downloadslreports/uncertainty communicationifiidfl RePii~NwS:E-2007-199, ISBN
978-90-8672-026-21 Copernicus Institute for Sustainable -~velom#~t and lnnov~bn, Departinent of Science Technology and Society (STS), Heidelberglaan 2, 3S~;p.l);.!J.~Ii~'~~sr.J'letherlaitds, Phone+ 31-30-2537631, Fax +31-30-2537601]
.. r
'::~:;::
~::::: .. -/
4 [Journal! Ke, J, Gong, T., Wang, WI Comml!l).i~pPns in <!;3inputation~jPhysics !"Language Change and Social Networks" I Page 935-9491 Vol. 3, No:''ij" ~~;:~~;;; ... /
5 [Journal! Ke, J., Yao, Y I Journal of Quantitative Li~i!~t:!_cs I "Analyzing Language Development from a Network Approach I Date of Publication: 20081 Page 70:~;:1:!\'RI, 15, No. 1]
6 [Labov 1 "The Social stratification of English in!:'~~~ Y od1~elty" 1 Date of Publication: 1966]
7 [Journal! Shelton, Donald E. I "The 'CSl Etfec.t:f~es It R~lly Exist" I NIJ Journal No. 2591 Date of Posting
17 March 20081 URL: www.ojp.usd9j.ppyfnijij6iirnals/259]
8 [Newspaper I Weiss, Rick I Wasbi~irt~~;Posti"CSI Effect Vexes Real Sleuths" I Date of Publication: 21
February 20051 Page A13]
::. /
. :
9 [Bulletin I Shane, Jon I FBl;a~~~fQre,~Qlili Bulletin I"Compstat Process" I Date of Publication: April20041
Page 12-21; and FBI La~;-~~forceri\tffi~~Biiitetin 1 "Compstat Design" 1 Date of Publication: May 2004 1 Page 12-
19]
.:~.~~=-=~~:-
/~~:~~;i~;...
.
10 [ Journ_!l:i~fsiiTiO.n~~A I"Rii~9P.Ill ~boice and the Structure of Environments" 1 Date of Publication: 19561
Psych:,y;~v. 63: P~te 129-13s''i1~$iffion HA. 19571 Models of Man. New York: Wiley]
11 [Bci6k:j;~}lllstc<iii'CR 1 Behavioral Law and Economics 1 Date of Publication: 2000 1 Page 3-91 Cambridge:
Cambridg~;Oniversity Press 1
12 [Taleb NN;:Q~te.ofP-ublication: 2008)
13 [Journal! Fi~H~i(B.; Lichtenstein, S; Slovic, PI Behavioral Decision Theory Perspectives on Risk and
Safety; Acta Psy_cllologica I Date of Publication: August 19841 Page 183-2031 Volume: 56, Issue: 1-3]
14 [Journal! Pidgeon NF; Poortinga WI Risk Analysis !"Trust, the Asymmetry Principle, and the Role of Prior
Beliefs" I Date of Publication: 2004124(6): Page 1475-1486]. 24(6):1475-1486]
IS [Book I Gigerenzer G, Selten R.I"The Adaptive Toolbox: In Bounded Rationality" l Date of Publication 2002
I Page 37-491 Cambridge, MA: MIT Press]
16 [Book I Kahneman & Tversky 1 Judgment Under Uncertainty: Heuristics and Bias I Date of Publication: 1982 I
Cambridge University Press]
17 (Book I Janis, I. & Mann I Pecision-Mal<ing: A Psychological Analvsis of Conflict. Choice. and Commitment
I Date of Publication: 19771 Free Press]
18 [Katz et aii"Network Theory and Small Groups" l Date of Publication: 2004]
19 [Book I Janis, Irving I Groupthink I Date of Publication: 19721 Houghton Miffiin]
20 [Book I Hart, Sundelius, & Stern eds.j Beyond Groypthjnk; I Date of Publication: 19771 University of
## Se~
Michigan Press]
21 [Klein, Gary; Hoffman, Robert R. eds.j"Using Knowledge Engineering to Preserve Corporate Memory" and
"'The Psychology of Expertise: Cognitive Research and Empirical AI" I Date of Publication: 1992 I New York:
Springer-Verlag]
22 [Book I Connell, M.; Gardner, H; Sheridan,K I The Psychology of Abilities, Competencies, and Expertise I On
"Abilities and Domains"! Edited by Robert Sternberg and Elena Grigorenko I Date of Publication: 2003 I
Cambridge University Press]
23 [Adapted from Chi 12006]
24 [Adapted from Clark !"Building Expertise~' I Page 9-15]
25 [Gvlaskeiwicz & Rulke !"Distribution of Knowledge, Group Network Structure, and Group Performance'1
26 [Journal! Cvetkovich GT; Siegrist M. 1 Risk Analysis !"Better Negative Than Positive? Evidence of a Bias for Negative Information About Possible Health Dangers" 1 Date of Publication: 2001 121 (I )::;P~ge 199-206]
27 [K~hlthau I "Inside the Search Process: Information Seeking from the Users' Perspective"]:;~:;:;~.
28 [Book I Preston, Thomas 1 The President and His Inner Circle: Leadership Style and the A~~j~~fifrocss in Foreign Affairs I Date of Publication: 2001 1 Columbia University Press]
. ::::;'".
:ii~~;::{
29 [Book I Boin, Hart, and McConnell! Crisis and After: The Politics oflnve5tigationG'i\~"C6iiiitabilitv;~~
kearning I Date of Publication: 20081 Cambridge University Press]
.-.
;:;~~:
:;::~:;.;:
30 [Website I Dynamic Analysis Spreadsheet I
:::;
.:'":;:;:;::..
;:
http://www.sharp.gov:8080/knowledgetree/view.php?tDocumentld=376
;:;:::.
::~:t:::
31 Measurement and Signals Intelligence, which includes Technical Nucl~-~~~l!ics ~-
32 [Journal! McDonnell, J. & Stumpf, S.i Journal of Universal Co!11PU:tlt"Scienc~ij~~gInvestigation into Sharing Metadata: "I'm not thinking what yo_u are thinking"! Datc:.ofPublj~t!qn: 20041 Page 740-748 1 Vol. I 0, No. 6]
33 [Book I McDonnell, J. & Stumpf, S.l Proceedines-oOhe~;i~lli Interii~ti!mal Worksbon on Database and Expert Systems APPlications CDEXAl 1 "Sharing Metada!;l!:'ProbleifiS;~and Poteifti~J;S~Iutions" 1 Date of Publication:
w~
-~~
~
7
34 [Website I Center for Strategic Intemationai Studl~~~;t.pa~e of Publication: 5 August 2009)
http://csis.org/prograrnlproject-nuclear-issues]
-::~::::~.
3s [H.R. 730 I "Nuclear Forensics and Attribution A~t"!26~:~hioo9]
.... ::;:
~ ..
~;~::