My particular specialty was reconnaissance intelligence, frequently referred to as "imagery" intelligence. My title changed from "photo interpreter" to "imagery analyst" as technologies changed, and more and more of my time was spent in interpreting non-visible-light imagery. That's just one specialty among a dozen or more:
- Human Intelligence (HUMINT) is the oldest form of intelligence, and is still important today. Human intelligence includes spies, moles, and agents, and the information these people can get to their "handlers". It also includes such commonplace activities as debriefing people who have visited certain areas, or who have had contact with certain people from "listed" countries.
- Signals Intelligence (SIGINT) includes intercepting and interpreting electronic signals. It can include everything from telephone to television to encrypted data transmissions. My mother was an enlisted signals analyst for the US Navy during World War II disciphering Japanese naval communications, a form of SIGINT.
- Electronics Intelligence (ELINT) concentrates on what part of the electronic spectrum is being used, rather than what the signals themselves might contain.
- Imagery Intelligence (IMINT) uses some form of recording system to acquire representative data (photographs) of surface features, including military, cultural, agricultural, economic, transportation, and other specific data.
- Communications Intelligence (COMINT) includes a bit of SIGINT and ELINT, but is primarily trying to discover how groups communicate (electronically, mechanically, etc.), as well as what they're saying, and when they're talking to one another.
Each of these types of intelligence (there are others, much more specific, and not commonly discussed) have dedicated people who spend years learning and perfecting their skills in their particular job, and who develop a degree of expertise in what they do. Yet their primary function is to report what they learn to others, who will interpret the information, compile the data, correlate the reports, and create summaries and in-depth studies.
Intelligence, regardless of kind, follows a cycle: targeting, collecting, evaluating, interpreting, analyzing, collating, and distributing the data. Intelligence personnel just don't go off into a bare spot and begin collecting data: the national intelligence community determines what kind of information the nation needs, and the best way(s) that information can be gained. More than two-thirds of the information needed by the US intelligence community can be gathered from "open sources" - data produced by nations, international organizations, scientific, technical, cultural and social publications, daily newspapers, and similar sources. Sometimes these sources will be questioned, and the task will be issued to verifiy or disprove a particular piece of information. Some data cannot be collected by such "open" means. That's when the national intelligence community is provided tasking to determine the information. Tasking takes into account several factors, including the most appropriate means of collecting the necessary information, how important the information is, the accessibility of the information, the usefulness of various sources, the parameters of the necessary data to satisfy the requirement, and the timeliness of the collecting/interpreting/reporting process.
Once the collection requirements are finalized, they're sent to the appropriate collecting agency (or agencies). The collecting agency does its best to satisfy the requirement, depending upon the priority of the data needed, other priorities, and collecting tools available. These collecting agencies do their best to satisfy the collection request. It may take several partial collections to collect the needed data, or several failed attempts before an attempt is successful.
Once the raw data is collected, someone (usually several someones, to ensure accuracy) gets the data to evaluate and interpret. There are many internal (and classified) factors that enter into the picture, including whether the data is collected at one time, or over time; whether the collection request has been partially/mostly/completely satisfied, whether the data is of a, b, ... x, y, or z quality (and if that quality is sufficient to meet the tasking), and so forth.
Someone, usually the analyst, then generates a raw report, based upon the data being interpreted. Most of the time, the report will be edited for accuracy, inclusiveness, comparisons, and many other factors, then transmitted to a list of authorized recipients. At this point, the report is still "raw" intelligence, although the analyst may have used any number of previous documents, plus his own extensive experience, to make the report.
Most major combat commands and even subordinate units today have what they refer to as a "fusion center". This is where raw reports such as the one mentioned above are sent. They are, in turn, read by people specially trained to combine reports from many different sources, compare those reports to the previously known information, and decide if what they're getting is a significant change, a minor change, or no change. These people analyze data from many different sources, compare it to what they already knew, and determine if there's something important that needs to be passed on. These are also the people that end up developing briefings for senior commanders and their staff, all the way up to the White House. These people also create periodic combined-intelligence reports covering a particular weapons system, area, operational capability, unit, or some other specific grouping.
"Intelligence", then, is the process of collecting, interpreting, analyzing, collating, reporting, and distributing useful information. It's not a static process - what's learned today may change tomorrow, or may continue for months or years. Intelligence is not only classified because of the information it might contain, but also the capabilities to collect or analyze that information - sources and methods. One of the chief areas that we want to keep secret is just how very good (or bad!) we might be at a given type of intelligence collecting, or about problems and limitations our collecting methods might include.
Needless to say, we (the Intelligence community) don't always tell everything we know, or how we got the information we do have. No one wishes to give the enemy knowledge they haven't earned, or that might be detrimental to US forces, either today or twenty years from now. That makes it very difficult to judge the effectiveness of our intelligence services - not only for those not included inside the community, but for our enemies as well. Claiming the Intelligence community isn't doing its part, or has "failed", is the whine of someone that doesn't have information they feel they should have, as some type of right. The success of our military forces in responding to our enemies is proof that our intelligence is very, very good. The claims of intelligence failures in Iraq are far more wishful thinking than actual fact.
Disclaimer: I do not have access to any classified information at this time, nor am I communicating with anyone in the US intelligence community ANYWHERE on a regular or frequent basis.