Does US government secrecy threaten national security?

Columbia University History Professor Matthew Connelly discusses his
new book on “The Declassification Engine: What History Reveals About
America’s Top Secrets” with Radio Active Magazine regular Spencer
Graves and guest interviewer Doug Samuelson.

Connelly’s History Lab at Columbia University has used text mining
and artificial intelligence tools in studying all the declassified US
government documents they can find. Among other things, they found
substantial documentation supporting the claims that the Roosevelt
administration had worked to maneuver the Japanese into “firing the
first shot”, which Roosevelt then implied was “unprovoked“.

Connelly drew “Three major lessons from Pearl Harbor”:

  1. “Secrets are often kept to hide incompetence.”(p. 55)
  2. “[C]atastrophic attacks almost never come out of nowhere. …
    [L]eaders prefer to pretend they had no warning. In fact, claiming that
    an attack is completely unexpected can help legitimate more espionage,
    domestic surveillance, and military spending.”(p. 55)
  3. “[O]ur leaders think we can’t handle the truth and wouldn’t support
    their plans if we knew what they were.”(p. 57)

Other documents describe how President Johnson in 1964 authorized US
support for attacks by South Vietnam against North Vietnam.  He did that
to provoke an incident that he could denounce as “unprovoked” to avoid
losing that year’s presidential election by appearing soft on
Communism.  Similarly, other sources imply that if President Clinton
had treated the 1998 bombings of US embassies in Kenya and Tanzania as
law enforcement issues, many involved in the preparations would likely
have been brought to justice, and the suicide mass murders of September
11, 2001, would likely not have occurred.

And many classified documents are destroyed in violation of US law to prevent them from ever
coming to light, thereby thwarting the ability of the historians and voters from learning from the past and using that to form a more perfect union.

Connelly suggests that artificial intelligence algorithms can be developed that would better protect the nation against unauthorized leaks while also providing the public with much better information.

Copyleft 2023 Matthew Connelly, Doug Samuelson, Spencer Graves Creative Commons Attribution-ShareAlike (CC BY-SA) 4.0 International license


Share This Episode