Friday, July 11, 2008

What Is Intelligence? (Final Version With Summary)


  • There is no standard definition of “intelligence”. Popular thinking and the best efforts of legislatures, agencies and academics to the contrary, no generally agreed upon definition of intelligence exists. This problem is exacerbated when the newly formed intelligence communities in law enforcement and the private sector are included.
  • Developing such a definition is important in order to create realistic expectations in the minds of the decisionmakers intelligence is designed to support. This is particularly true in a democracy where the electorate views the notions of secrecy and unaccountable power often linked with intelligence activities with hesitation.
  • Two activities, secrecy and covert operations, typically associated with intelligence are not, in fact, necessary to define intelligence. Secrecy, or more accurately, confidentiality, is only necessary to preserve options for the decisionmaker that the intelligence activity supports. Covert operations, on the other hand, are better viewed as an act of policy than as an intelligence activity.
  • Common threads run through many of the earlier attempts to define intelligence, however. These threads, pulled together, result in a good working definition of intelligence:

    • Intelligence is a process, using primarily unstructured information from all sources and focused externally, that is designed to reduce the level of uncertainty for a decisionmaker.

Online Version:

Part 1 -- The Problem Of "Intelligence"
Part 2 -- The Importance Of A Clear Definition Of Intelligence
Part 3 -- The Reasons For A Lack Of A Definition
Part 4 -- What Would A Good Definition Look Like?
Part 5 -- Previous Attempts To Define Intelligence/Legislative Attempts
Part 6 -- Previous Attempts To Define Intelligence/Agency Attempts
Part 7 -- Previous Attempts To Define Intelligence/Expert Attempts
Part 8 -- Previous Attempts To Define Intelligence/Law Enforcement And Private Sector Attempts
Part 9 -- Defining Intelligence


crussell82 said...


Thank you! You wrote an excellent article. I like and accept your working definition of intelligence. In a purely academic environment, it's nice to spend time thinking about how to define intelligence. In the bigger picture, it's not a need but rather a want. As humans, we always feel the need to define things and put them in a box. For example, you could spend all the time you want to make a working definition of love, but either way, everyone will have their own definition based on their own experiences. It's pretty much the same with intelligence. You can spend all the time you want to define intelligence, but everyone involved with intelligence will define it based on their personal experiences. Even if a legal definition were enacted by law, very little would change at the workerbee level doing their daily intelligence duties.

As a caveman, I would shorten your working definition to "Intelligence is processed information for a customer."

The purpose of intelligence is for actionable warning. Actionable infers that the intelligence was timely, accurate, and relevant. Additionally, if a customer decides not to act on intelligence that is by definition an action.

Again, I really enjoyed your article. Extremely insightful!!!


Kent Clizbe said...


You've done rigorous academic work that provides us with a starting point to defining intelligence.

"Intelligence is a process, using primarily unstructured information from all sources and focused externally, that is designed to reduce the level of uncertainty for a decisionmaker."

So, let's apply the proposed definition to real-world scenarios and attempt to differentiate what is, and what isn't Intelligence.

Scenario 1:
My 12 year-old son is trying to decide which character to add to his tabletop Lord of the Rings game. He consults a paper catalog, visits a website with a discussion group that has threads about the characters of interest, logs into and participates in a chat room where people are knowedgeable about the game, telephones a buddy to discuss the merits of Balrog, goes to a hobby store and gathers information about prices and features by examining the actual figures, engages the clerk in conversation about the materials used in making the figures, as well as numerous other information-collecting activities. After he has the information he has gathered from external sources, he sits down and analyzes it. He decides to spend $25 on the Mines of Moria set.

Scenario 2:
The Chief Operating Officer (COO) of a software company needs hire a Director of Taxonomies (DirTaxon) because the company plans to expand their online advertising services, and the COO has determined that a Taxonomy group will be needed to properly structure the concepts and assist customers in finding appropriate solutions.
The COO contacts his Human Resources (HR) director and directs her to find a DirTaxon, and lets her know that it was needed yesterday. The HR group immediately does a search on keywords through their in-house resume database. When that doesn't turn anything up, the HR specialists search through online resume banks. Coming up with nothing again, the HR director turns to a third-party recruiter. The third-party recruiter knows of a headhunter who specializes in Taxonomy and related fields. The first headhunter makes a split-fee arrangement with the second headhunter. The Taxonomy-speciality headhunter calls and emails several Taxonomy specialist who he knows, letting them know that there is an opportunity in Taxonomy. George S. receives the email, and though he isn't interested, forwards it to Sarah M. Sarah responds to the Taxonomy headhunter, and sends her resume. The Taxonomy headhunter calls her, and after several lengthy phone conversations, determines that she is ready, willing and able to do the job that the COO requires.
The Taxonomy Headhunter presents Sarah as a candidate to the COO. The COO calls Sarah, and after a long phone interview, invites her to the company for personal interviews. After three visits, meeting various team and management members, the COO makes an offer, and Sarah decides to accept. They negotiate a salary and starting date.

Now, let's run these scenarios through the definition of Intelligence, which has 5 critical attributes:

1. A process
2. Primarily unstructured info
3. Info from all sources
4. Info is focused externally
5. Designed to reduce uncertainty for a decisionmaker

Scenario 1:
1. Yes, my son engaged in a process
2. Yes, he used primarily unstructured info
3. Yes, the info was from "all sources," or at least a variety of sources
4. Yes, the info was focused externally
5. Yes, the process reduced uncertainty for a decision-maker

So, with 5 "yes" answers, we learn that my 12 year-old son is an intelligence practitioner!

Scenario 2:
1. Yes, the organization, the COO, HR, the headhunter, and Sarah all engaged in a process
2. Yes, the process used primarily unstructured info
3. Yes, the info was from all sources
4. Yes, the info was focused externally
5. Yes, the process reduced uncertainty for decision-maker(s)

So, with 5 "yes" answers, we learn that the process of a company realizing that it needs a specialist and then filling that identified need, is Intelligence. In fact, we could run each of the actors in this scenario, the COO, HR, the headhunters, even George and Sarah, through the definition, and get the same result. They are all Intelligence Practitioners!

It would seem that neither of these scenarios would be considered Intelligence by either Intelligence professionals, the participants in the process, or the vast majority of the public.

In sum, defining the Scenarios above as "Intelligence" is counter-intuitive.

There are numerous other examples from every-day personal and business life which could be run through the working definition of Intelligence, and provide counter-intuitive results (a PhD candidate's process, the process of buying a car, the process of choosing a spouse, etc etc).

I submit that using a definition of Intelligence that is so open that it includes a kid buying an action figure, is not only counter-intuitive, but counter-productive to the profession of Intelligence.

I submit that we need to work on definitions of separate Intelligence Disciplines, and not on an all-encompassing definition that includes so much that it becomes meaningless.

A more productive path might be to define:

1. Intelligence Collection
2. Intelligence Analysis
3. Intelligence Dissemination

I also submit that the word Intelligence has connotations of secrecy, espionage, and covert action. This connotation cannot simply be ignored out of existence. Picking and choosing among published definitions or creating new definitions that ignore this connotation are counter-productive.

As an analogy, consider the definition of "Doctor." We could broaden the definition of Doctor to: "Someone who takes an interest in the human body, examines bodily systems, and takes action on the findings of those examinations."

That results in the same counter-intuitive results as above. A 12 year old boy who discovers a splinter in his finger and removes it meets the definition. But he would not/not actually be a Doctor.

I further submit that the proposed definition of Intelligence is not/not a definition of Intelligence, but rather a definition of "Decision-support Science."


Kent Clizbe
571 217 0714

Charlie said...


Thanks for the very kind words. I do want to note, however, that I did not think of this as an ivory tower exercise. "Intelligence" in my mind is more like "law" or "engineering" than it is like "love" or "joy" or "sadness". As such, it should have a more rigorous definition to it. If intelligence is ever going to be more of a profession than a craft, such basic theoretical building blocks seem essential.

I also have to point out that your synthesis of my definition leaves out two crucial aspects, external focus and reduction of uncertainty. In short, your condensed version could be applied to a number of activities other than intelligence. This, in turn, makes it less useful as it fails to separate intelligence activities from other activities.

I don't mean to be argumentative here. I actually very much appreciate that you have taken the time to comment and I am more than happy to continue the dialogue.


Charlie said...


Wow! Where do I start? Thanks for the detailed response. I will try to do it justice.

First, with regard to your 12 year old son. Yes, I agree that he is, among other things, doing intelligence. It is a trivial case (though he might disagree) but I see no real difference between this and the examples I used to counter the argument of "importance" in part 2 or 3 (I forget which). When you build a doghouse you can be said to be doing engineering and architecture. The mere unimportance of the activity does not remove it from the "big tent" definition of these activities and unimportance should not remove intelligence like activities from the big tent of intelligence. In fact, most hard science disciplines use these simple examples as thought experiments on which to build theoretical concepts. My precise intent was to create a big tent definition.

In your second example there are a number of things going on to support the decisionmking process and I agree that not all of them are intelligence. Searching in house resume databases are not intel in that they are not externally focused, for example. In addition, the external search does not seem to really be looking at all sources of information but merely those sources that are traditionally associated with the hiring process. I may come across as quibbling here but it is pretty clear that some of these things are intelligence and some are not and the mere fact that they are "normal" activities for HR professionals does not make it any less intelligence than using a hammer to drive a nail is a normal architectural activity.

The example also raises a much more interesting question. Can intelligence support more than operations?

We don't normally think of it that way but I know of many personnel and logistics decisionmakers who bemoan the fact that intel traditionally does such a poor job of supporting them since it spends so much time supporting ops. I don't know why, from a theoretical standpoint, intel can't support personnel. Maybe the issues with the personnel example are more about our biases with respect to intel than the way intel really is.

To take the point a bit further, let me suggest another example. Lets say your son goes through his LOTR figures and decides, based on a wide variety of internal reasons, to play with just six of them. Clearly there was a good bit of decision support activity going on as he figured out which ones to play with and why. It is also very different from the kind of activity where he went out and researched externally about which set of figures to buy. Whatever you want to call it, there seem to be at least two different kinds of decision support activities -- one that focuses primarily on internal resources and one that actively gathers information from external resources and about external actors. Just as obviously, the best decisions require both.

I do have to object to the idea that I ignored secrecy out of existence. I merely said that it was not necessary for a definition of intelligence. I went on further to say (in my rather detailed discussion of this topic in the posts themselves) to talk about the role of secrecy in preserving decisionmaker options. Adopting this path means that secrecy for secrecy's sake cannot be the part of any legitimate definition of intel -- it is subordinate to the decisionmaker's requirements.

Finally, and I realize I am out of order here, I can't see how we can "work on definitions of separate Intelligence Disciplines" without understanding first what intelligence is. I suspect I am missing something but it looks on its face as something that is impossible.

Again, my intent here is not to be argumentative but to engage in the dialogue. If it comes across as abrupt or in any way dismissive, you have my apologies.


crussell82 said...


Continuing the dialogue. The three remaining points are probably where our thoughts differ, which is fine: ) Firstly, I don’t consider “Intelligence” to be like law or engineering. The law is a series of interpretation based on historical precedence. Additionally, there is a level of authority (judge/judges) that can make a final decision. Intelligence doesn’t have neither historical precedence nor a final decision authority. With regard to engineering, other than the fact that certain intelligence disciplines follow scientific laws as in engineering such as SIGINT, IMINT, and MASINT technologies, the actual intelligence analysis is more of a human endeavor rather than a scientific method. However, as is human nature, we try our best to create structured analytic techniques and call it scientific to make ourselves feel more legitimate.
Bottom line: Intelligence in its basic form is trying to predict the intent/actions of another person or entity, which is more of an art than science.

Secondly, I may be confused by what you mean by “external focus.” If you mean, outside of the U.S. or U.S. persons/entities, then again, I have different thoughts. In the past, Intelligence was externally focused by law and organization of the Intelligence Community (IC). However, since 9/11, the IC now stresses the importance of both the FBI and DHS. Both of these agencies by law have an internal focus with regard to U.S. persons/entities. Both continue to grow their Intelligence responsibilities, roles, and missions. Additionally, FISA was setup to allow for legal internal collection in the U.S. and on U.S. persons/entities with relation to terrorism.

Thirdly, with respect to reduction of certainty, much of the intelligence production has little to do with reduction of certainty. For example, as data is collected, then exploited, many of the products are put into databases that while are available to customers are rarely seen by anyone because it’s not a priority at the time. Africa is a prime example of this. Additionally, even if you look at such important products such as a National Intelligence Estimate (NIE), I would argue that their vagueness tends to create opportunities for different interpretations depending on what the customer wants to support (ie Iraq WMD, Iran nukes, etc). If anything, Intelligence products at times increase the lack of certainty.

Again, thanks for the great work you're doing. Thank you for listening to me babble: )


Kristan J. Wheaton said...


My point was that "intelligence" is more like law or engineering than it is like "love". I agree that there are many distinct (and obvious) differences but, as I used them in the series of posts, I felt like there were some strong commonalities as well. They all purport to be professions and, more importantly, they are applied professions -- professions where the worker does something more or less directly for someone. This distinguishes (in one way) these activities from the fine arts and some scientific endeavors. Finally, as a concept, in each case, they can, indeed they must, extend themsleves beyond the realm of the professionals who practice them. A kid builds a bridge out of legos and it collapses when he runs his toy truck over it. Engineering principles can explain why, even though the child is not an "engineering professional". This is the big tent definition of intelligence I was trying to get at.

I agree that the concept of external focus can be tough. I was trying to use shorthand here in order to make a more concise definition. I laid out the full argument for that phrase in detail in the posts but the quick version is that intelligence is primarily about "the other guy" -- the criminal, the competition or the other country/transnational group. This really distinguishes intel from most other decision support activities. It doesn't mean that people doing intel dont need to know anything about their own organization but that the primary focus is on the other guy.

With respect to uncertainty, I certainly agree that there are a number of documents produced that either sit unread or fail to reduce uncertainty. I think though, that theoretically, the purpose behind their creation was, in fact, to reduce the uncertainty facing a decisionmaker in some, perhaps minor, way. For example, one of the purposes of laws is to serve justice. It does not always happen that way but setting such a goal as the theoretical ideal allows legal professionals to know what the system should do and adjust it so that future results can better conform with that goal.

Thanks for commenting again!


PS sorry about the previous replies being posted as "Charlie". I was unwittingly using my son's account...

crussell82 said...


I really enjoy the exchange of ideas. Thanks for the opportunity. I assumed you were using the name "Charlie" as denial and deception:) I've bookmarked your blog.

Very respectfully,