More than a decade after the 9/11 terrorist attacks, a set of extraordinary and secretive surveillance programs conducted by the National Security Agency has been institutionalized, and they have grown. 

These special programs are conducted under the code name Ragtime, and are divided into several subcomponents, according to the new book Deep State: Inside the Government Secrecy Industry, by Marc Ambinder and D.B. Grady. (I purchased a copy this morning.) 

The authors, both journalists who cowrote a previous book about special operations in the military, have dug deep into the code names and operational nitty gritty of the NSA's secretive and hugely controversial surveillance programs, and they've come up with impressive new details. 

Ragtime, which appears in official reports by the abbreviation RT, consists of four parts. 

Ragtime-A involves US-based interception of all foreign-to-foreign counterterrorism-related data;  

Ragtime-B deals with data from foreign governments that transits through the US; 

Ragtime-C deals with counterproliferation actvities; 

and then there's Ragtime-P, which will probably be of greatest interest to those who continue to demand more information from the NSA about what it does in the United States. 

P stands for Patriot Act. Ragtime-P is the remnant of the original President’s Surveillance Program, the name given to so-called "warrantless wiretapping" activities after 9/11, in which one end of a phone call or an e-mail terminated inside the United States. That collection has since been brought under law, but civil liberties groups, journalists, and legal scholars continue to seek more information about what it entailed, who was targeted, and what authorities exist today for domestic intelligence-gathering. 

Deep State has some answers.  

Only about three dozen NSA officials have access to Ragtime's intercept data on domestic counter-terrorism collection. That's a tiny handful of the agency's workforce, which has been pegged at about 30,000 people.  

As many as 50 companies have provided data to this domestic collection program, the authors report. 

If the NSA wants to collect information on a specific target, it needs one additional piece of evidence besides its own "link-analysis" protocols, a computerized analysis that assigns probability scores to each potential target. This is essentially a way to use a computer data-mining program to help determine whether someone is a national security threat. But the authors find that this isn't sufficient if NSA wants to collect on said target. And while the authors found that the Foreign Intelligence Surveillance Court rarely rejects Ragtime-P requests, it often asks the NSA to provide more information before approving them.   

How the surveillance is approved tells us a lot about the breadth of the NSA's intelligence gathering. The court and the Attorney General both certify a slate of approved targets under Ragtime-P, the authors find. That includes a certain amount of "bulk data"—such as phone call logs and records—that can be collected around those targets. An NSA official told the authors that Ragtime-P can process as many as 50 different data sets at one time. 

What happens next looks like a 21st-century data assembly line. At the NSA's headquarters in Fort Meade, Maryland, a program called Xkeyscore processes all intercepted electronic signals before sending them to different "production lines" that deal with specific issues. Here, we find another array of code names. 

Pinwale is the main NSA database for recorded signals intercepts, the authors report. Within it, there are various keyword compartments, which the NSA calls "selectors." 

Metadata (things like the "To" and "From" field on an e-mail) is stored in a database called Marina. It generally stays there for five years. 

In a database called Maui there is "finished reporting," the transcripts and analysis of calls. (Metadata never goes here, the authors found.) 

As all this is happening, there are...REST OF IT