from the FAS Project on Government Secrecy
Volume 2014, Issue No. 65
October 1, 2014
Secrecy News Blog: http://fas.org/blogs/secrecy/
- NUCLEAR ATTRIBUTION AND HOT COGNITION
- CIA ASKS TO DESTROY EMAIL OF NON-SENIOR AGENCY OFFICIALS
- CIA SAYS REDACTION OF COMMODORE AMIGA COST WAS AN ERROR
NUCLEAR ATTRIBUTION AND HOT COGNITION
The challenges of identifying the perpetrators of a nuclear attack on the United States and communicating that information to senior leadership were considered in a 2009 workshop sponsored by the Office of the Director of National Intelligence. A declassified report on the workshop was released last week in heavily redacted form.
See "Transforming Nuclear Attribution: Culture, Community, and Change (SHARP 2009)" (redacted), Office of the Director of National Intelligence, July 2009:
One of the challenges is that the task may be impossible. "The outcome from the assessment of all the evidence and sources may be that a definitive answer is not achievable."
In the best of cases, "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 attributing responsibility for an event)."
"Given the magnitude of the likely national response to any substantial WMD event, those involved in the attribution process need to be cautious of leaping to conclusions ahead of the evidence."
The report considers the problem of "hot cognition," referring to analysis that is performed under conditions of emotional agitation or distress.
"Hot cognition has an immense potential for distorting our perceptions of the environment and how we interpret information. It leads us to more extreme judgments of information, perhaps far beyond what they warrant. And it may lead us to fill in the gaps of missing or ambiguous information with emotional filler that could seriously distort our assessments."
The workshop was conducted as part of the ODNI Summer Hard Problem (SHARP) program. In the roughly 50% of the resulting report that was not redacted by ODNI and the Department of Energy, there are a number of passages of interest concerning the psychology of intelligence analysis, and other topics. For example:
* "Resolving an information need is not just about finding a 'nugget.' Information must be actively incorporated into the mind. New information has to be assimilated into a person's preexisting 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."
* Assembling an "all-star" team of outstanding intelligence analysts to tackle the attribution problem may not be the right approach. "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 atheletes individually, these all-star teams typically do not perform as well as expected, nor do individual all-stars perform as well as they performed on their originating team."
* "In the case of a nuclear event, it is likely that individual private citizens will have images stored on cell phones or digital cameras that could help [resolve] the attribution question. [Word deleted - AFTAC?] should make arrangements in advance of any actual emergency that would give the public a way to send information to government servers for analysis."
* "For every one casualty actually caused by a [WMD] event, as many as fifty other individuals may descend upon local medical facilities presenting with psychosomatic symptoms."
* "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 cannot be allowed to impede the higher priority of protecting thousands or millions of human lives.... We must prepare IT tools and approaches now, that when activated for a nuclear emergency, allow relevant players to share knowledge at the speed of technology, not the speed of bureaucracy."
* "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 [intelligence community, law enforcement, and technical nuclear forensics] 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*."
CIA ASKS TO DESTROY EMAIL OF NON-SENIOR AGENCY OFFICIALS
The Central Intelligence Agency has asked for authority to destroy email messages sent by non-senior officials of the Agency. The National Archives and Records Administration (NARA) has tentatively approved the proposal.
In an August 18 appraisal of the CIA request, Meredith Scheiber of NARA wrote that any permanently valuable material in the emails would almost certainly be captured in other permanent CIA records.
"It is unlikely that permanent records will be found in these email accounts that is not filed in other appropriate files appraised as permanent," the appraisal said.
"There are multiple records systems to capture the actions and decisions of employees and multiple internal controls in place in the event an employee was engaged in malicious activities."
Any "remaining email not captured in other recordkeeping systems is routine or administrative in nature; transitory; or personal in nature."
The NARA appraisal of the CIA proposal noted in passing that "The Agency's current email policy is to print and file" rather than to save permanently valuable email in softcopy format.
"The average career of an Agency employee is 22 years," the NARA appraisal also observed.
The CIA proposal for email disposal authority and the accompanying NARA appraisal were announced for public comment in the Federal Register on September 17.
CIA SAYS REDACTION OF COMMODORE AMIGA COST WAS AN ERROR
The CIA should not have redacted the amount that was paid for a Commodore Amiga portable computer in 1987 from a recently declassified article, a CIA official said today. ("CIA: Cost of Personal Computer in 1987 is a Secret," Secrecy News, September 29).
"The redaction of the cost of the Commodore Amiga computer was in fact an error," said Joseph W. Lambert, Director of CIA Information Management Services.
"Although we would normally redact budget figures, this clearly does not constitute a budget figure and should not have been redacted. The mistake was made in a high volume court deadline environment," he said, referring to a FOIA lawsuit brought by former CIA official Jeffrey Scudder.
"I have instructed my folks to make the appropriate corrections by lifting the redactions in question and then subsequently re-post the document to our website," Mr. Lambert said via email. The revised document should be posted tomorrow.
The Scudder lawsuit was not settled by the latest releases of hundreds of articles from CIA's Studies in Intelligence journal. The parties told the court on Monday that Scudder intends to challenge some of CIA's withholdings.
Secrecy News is written by Steven Aftergood and published by the Federation of American Scientists.
The Secrecy News blog is at:
To SUBSCRIBE to Secrecy News, go to:
To UNSUBSCRIBE, go to:
OR email your request to email@example.com
Secrecy News is archived at:SUPPORT the FAS Project on Government Secrecy with a donation here: