When an agent or operative turns in summaries or documents, these are poured into the main intelligence files and any "special bank" that is of current interest.
These main files or "special banks" can be made to operate like computers without the bother of electronic liabilities.
Modified only by errors of concentrated attention where none should be going or omissions of areas that should be under collection, these files or "special banks" will divulge new data.
Here is where we leave the subject of intelligence reports (See HCO PL 25 April 1968, _INTELLIGENCE_ACTIONS_) and enter the field of ESTIMATION AND PREDICTION.
A main file is where everything is filed. A "special bank" is a separate file where data is ADDITIONALLY AND ALSO filed relating to an exact subject.
When filing and cross-indexing is kept up to date (and it always must be) it is not possible to begin the work of ESTIMATION.
This is done by examining the main files and locating the fattest folders related to any one group or subject. Let us say one is just now interested in dog catchers. The cross-file index, if properly prepared, will show that several dog catchers are in the main files. One simply pulls their files from the main file, leaving a dummy to catch new filings and to show where the main file is, and makes a "special bank". This is not the Dog Catcher Special Bank.
Now if you look at the above paragraph again you will see that SOMEONE had to say that "we are now interested in dog catchers." This in essence is an ESTIMATION in a crude form. One noticed by general observation that there simply seemed to be a log about dog catchers lately getting into the main file. So the head of the section being alert decides that there might be something growling in that area. So he rapidly makes up a "special bank" of all files that seem to be concerned to some degree with dog catching.
Noticing that we should be looking at these dog catchers is what could be called "flair". It's a "Hey, looking this over we seem to have a lot of dog catchers lately, so I better get some data on them and make an estimation."
No file or computer can actually THINK. It can however be so arranged that QUANTITY begins to show up around certain names and that a casual examination of those names shows they seem to have dog catching in common.
So an Estimator, working the files routinely, having noticed this, would make a "special bank".
The "special bank" is not very elaborate. It can be as crude as simply spreading out the folders on a desk and looking them over FOR THINGS THEY HAVE IN COMMON OR NAMES IN COMMON TO THEM.
It can be as elaborate as making a new card index for every name in those files and what file they came out of; then filing these cards in alphabetical order. As one has taken EVERY name out of each file in turn and put each on a separate card, it is obvious that when one then puts them in alphabetical order, the names in common to two or more files will occur as two or more cards.
If agents are trained to collect all names and connections of names on every investigation and if these names are supplemented by adding in directories, and if filing has been kept up so that a file exists for every name, when you choose files for a special bank they will be the fattest on that subject. And when you work the special bank over again, you will get the precise common denominators of people that run through several files.
You can now say, "Get me more data on Mr. Plow, Mr. Row, Mr. Bow and Mrs. Snap" and when that data comes in (and is added to main files but especially to this "special bank") you will suddenly find yourself expanding ABOVE the level that your special bank was at before. In other words, you have found new horizons.
You will find if you do this that Mr. Plow and Mr. Row are just goofballs or dogmeat sellers, that Mr. Bow died last year and that Mrs. Snap is all you have left.
So we look up Mrs. Snap even further, "Hey, go look over Mrs. Snap again" and we suddenly find she was receiving money for a fund called "Kennel Support" but was actually the head of a dog catcher ring.
We can now PREDICT.
If we are out to handle the dog catchers we will have to work on Mrs. Snap to get her to reform.
We can write up a program like this:
PREDICTION: DOGS WILL CONTINUE TO DISAPPEAR.
"The situation is that dogs are getting caught." "The WHO is Mrs. Snap." "The following program based on best advices should be carried out:
"1. Confront Mrs. Snap and get her attitude ......... etc."
One needn't elaborate for it is obvious that many ways exist to legally and ethically persuade Mrs. Snap to give up her career of crime.
The tools one is using consist of looking for people or groups that are themselves outpoints. This is about all the relationship this has to the Data Series.
The police use a crude version of filing which does not bear much resemblance to this. They have a RECORD of someone and they just keep adding to it and adding to it, putting in false reports, opinions, anything. If someone is "after" that person for some reason or another, they pull out the file and use it. This is a "dossier system" in such vogue with the Czar's Cheka
There is no excuse for a 'dossier system" as it has a suppressive action on everyone: the files one by one are used as a sort of punishment system. It exists to catch individuals without relation to general situations. Thus police continually find themselves up against increasing crime since they do not handle broad situations relating to crime. The reason this is apparent is that if they did use a different approach they would have found a long while ago that their violent crime was occurring amongst people recently in mental institutions or prisons and it would have led them to condemn the ineffectualness of the psychiatrist, psychologist and look for other means.
An exception is where an intelligence action is undertaken by police to combat organized crime or groups but here again they do not achieve reduction (as witnessed by statistics) as their system never leads them outside the group being investigated. They consume countless hours of "law enforcement" policeman and judge time which could be saved if they adopted a more streamlined system which, amongst other things, weeded out false reports or vexatious complaints with no foundation in fact.
The only point here is that filing systems DO exist other than the one being described and that the effectiveness of a filing system is rewarded by saving time and energy and the achievement of its actual purpose of situation predicting and handling.
One can vary this system of intelligence filing and making "special banks" in a number of ways. One can shortcut everything in it but the actual making of the main files and the actual pulling of "special banks" and studying them.
One can keep a main system of individuals flanked by a cross-index system of groups which shows how many individuals in the main bank fit into each specific group.
One can keep a main file of groups and flank it by "special banks" of individuals who are common to one another or to groups.
The whole point is that ALL data by name and separately by group must get into the main files.
The action of estimation is helped by a system that shows some GROUP getting fat with names.
The action of prediction is always done by forming a "special bank" for study and sorting out.
The action of programming comes from any data in the individual or the group file.
The material is from newspaper accounts, literature issued, corporation filings, from almost innumerable types of sources.
Knowledge of the Data Series outpoints prevents one from buying false reports in the files and careful cross-filing eliminates the innocent bystander; these two things can defeat any intelligence system: accepting false data and hitting innocent bystanders; this is the fault of nearly every system in existence.
Intelligence filing done well, estimated well, furnishes predictions which can be programmed and situations can be handled that would otherwise drop on one from an apparently clear sky.