Thursday, May 26, 2011

Part 5 -- Critiques Of The Cycle: Which Intelligence Cycle? (Let's Kill The Intelligence Cycle)

Part 1 -- Let's Kill The Intelligence Cycle
Part 2 -- "We''ll Return To Our Regularly Scheduled Programming In Just A Minute..."
Part 3 -- The Disconnect Between Theory And Practice
Part 4 -- The "Traditional" Intelligence Cycle And Its History

Despite its popularity, the intelligence cycle is widely criticized by intelligence professionals. These criticisms generally break down along three lines. First, there are those who say that what appears to be a theoretical monolith is actually open to a wide variety of interpretations, depending on perspective. Indeed, there is not one intelligence cycle but a series of intelligence cycles, each substantively different from the rest.

Second, many authors have claimed, quite convincingly, that the intelligence cycle, as generally described, does not, in many material ways, reflect the reality of how intelligence actually is done. The simplicity of the cycle, to these critics, is both seductive and deceiving.

Finally, there are those critics who claim that the so-called intelligence cycle is simply a marketing tool that rebrands overly simplistic "cycles" from business and leadership courses. I intend to discuss each in turn.

Which Intelligence Cycle?

Compare the diagram of the intelligence cycle which, until recently, graced the website (owned by the Director of National Intelligence -- DNI) with the diagram of the intelligence cycle from the Federal Bureau Of Investigation (FBI) below.

FBI Version Of The Intelligence Cycle

Recent DNI Version Of The Intelligence Cycle

Besides the obvious differences in graphic representation, what differences in content do you notice? If you look carefully, you will see that the FBI has decided to include a phase that is not in the DNI’s image, the “requirements phase”.

For a seasoned professional, this difference is trivial. Indeed, as I discussed in my overview of the intelligence cycle in Part 4 of this series, there is an explicit need for requirements and the FBI’s inclusion of them as a separate phase of the process might seem to be a matter of professional choice.

A student of intelligence, particularly a new student, might legitimately question this explanation, however. Perhaps there is a difference. Perhaps the FBI’s characterization represents a new way of thinking about intelligence as a process.

Perhaps, in fact, one description of the process is substantively better than the other. If this is not the case, then what is the explanation for the differences? There does not seem to be a good reason why the FBI’s take on the intelligence cycle should differ from that of the main intelligence site for the US Government, particularly since the FBI’s intelligence function, since the 2004 restructuring of the US intelligence community, is, in many ways, subordinate to that of the Director of National Intelligence. In short, does this difference represent legitimate theoretical differences or is it merely the result of a lack of coordination or, worse, sloppy thinking?

To make matters worse, the DNI's recently updated version of the intelligence cycle confuses the issue even more.  You can see the the graphic currently in use at below:

A quick examination of the current version of the DNI's cycle seems to differ from the previous version is several substantive ways.  "Direction", "Exploitation" and "Production" all appear to be subsumed into broader categories of activities.  Is there a reason for this?  Did the DNI conduct studies to determine the best, most accurate, description of the cycle?  Or was this a graphic design decision made becasue there was simply not enough room in the graphic for the additional words? 

It gets worse.

On the same page that contains the graphic above, the DNI promotes not one but two additional variations of the cycle.  In the first, more modest, variation (contained in the text that describes the picture), the DNI says, "The process begins with identifying the issues in which policy makers are interested and defining the answers they need to make educated decisions regarding those issues.  We then lay out a plan for acquiring that infromation and go about collecting it."  If this is true, then why doesn't the graphic also "begin" with requirements?  Why does the graphic seem to begin with planning?

It gets even worse.

The third variation of the cycle (all on the same page) from the DNI comes at the very top of the page.  Here one finds five links, "Management", "Data Gathering", "Interpretation", "Analysis and Reporting", and "Distribution".  Clicking on the "Management" link indicates that management -- not requirements, not planning -- "is the initial stage of the intelligence cycle".


I wonder which version is taught in the Intel 101 courses?

I wonder how you grade a student who uses an "alternative" cycle as an answer on a test?

I wonder, if the intelligence cycle is perfect (as about 15% of the people I have polled indicate), which of these cycles is perfect-est?

Were these differences the only differences within the US national security intelligence community, they might be explained away more simply but they are not. In fact, there is very little consistency across and even within a number of important elements of the US national security community. These inconsistencies also exist across disciplines as well.

Examine the chart below. Only one function, collection, is universally attributed to intelligence across all 10 organizations examined.

Within the DNI, CIA and FBI there are minor but important differences – not one of the three is exactly like either of the other two.

Even more baffling are the differences within the US military, however. The Defense Technical Information Center (“the premier provider of Department Of Defense technical information”) has a streamlined four-part description of the cycle, one that largely (but not completely) agrees with the cycle as taught at Fort Huachuca, the Army’s home for its military intelligence professionals. This cycle, however, is substantially different from the process defined in the US Military’s highest-level publication on intelligence doctrine, Joint Publication 2.0. 

The differences evident in the US military may well be due to different publation dates or my own lack of access to the most recent revisions of some of these documents.  In this regard, though, the 2007 Joint Pub is worthy of further commentary.  In it, the US military seems to abandon the intelligence cycle in favor of a more generic intelligence "process".  Some have suggested that this proves the military has already killed the intelligence cycle (but it just didn't get the memo...).

While it is (from my viewpoint, at least) a step in the right direction, it only exacerbates the impression that either the left hand is not speaking to the right in the US national security intelligence community or that the DNI doesn't control or doesn't care what the Joint Staff puts out with respect to the intelligence process.  All of those alternatives make the US IC look sloppy and disorganized.

I also think the Joint Staff is trying to have its cake and eat it, too.  Compare the two images below.  The first is from the most recent public version of Joint Pub 2.0.  The second is from the 1990 version of the US Army's Field Manual 34-3, Intelligence Analysis.  While the words in the two publications contain many significant differences, the pictures seem to say that the military has not backed too far away from its conception of the process as a cycle.

Join Pub 2.0 Intel Process 2007
FM 34-3 Intel Cycle 1990

These descriptions of the cycle differ, again, in significant ways from the descriptions provided by two oversight bodies commissioned to examine intelligence activities listed on the chart, the 1996 Graham Rudman Commission and the 2004 Weapons of Mass Destruction Commission. To round out the confusion, the description of the cycle offered by the International Association Of Law Enforcement Intelligence Analysts and the classic competitive intelligence model (as described by longtime private sector intelligence specialist, John McGonagle) also differ from each other and from the other 8 examples.

This analysis, while interesting, comes across as perhaps a bit more picky than it should. Other processes in other disciplines lend themselves to various descriptions. Indeed, despite the differences, there are clear themes that emerge even from this analysis. Few, for example, would question whether requirements, needs, direction, and planning fell into a single, generic category.

Themes, however, is all these are. A rigid approach to intelligence, implied visually in the pictures above and in many of the descriptions of these processes by each of these intelligence organizations, seems inappropriate under these conditions for teaching these concepts to new members of the intelligence profession or, indeed, explaining the process to the decisionmakers that intelligence supports. Instead, a more nuanced and less absolutist approach appears to be called for.

There is one specific area where this analysis does create cause for concern, however. Only three of the 10 organizations examined include a feedback or evaluation function within their versions of the cycle.

While some of the other organizations did include feedback as a subset of the dissemination process, subordinating this crucial evaluative process is not likely to endear the intelligence function to the decisionmakers that intelligence supports. It seems much better practice to include explicitly the role of feedback in the process, whether the decisionmaker chooses to take advantage of it or not.

Next:  The Intelligence Cycle vs. Reality


Dr. Harry Nimon said...

I have often been drawn to a comment I made when a shave-tail LT at MIOBC and being introduced to the 'cycle.' I asked the instructor why the cycle appeared to be linear and not multifaceted. The answer from this Major was, "Huh?" and that was all.

We all realize that we do not take a single requirement, plan it out for collection, obtain the data, analyze the data, report on its existance, then go out for the next bit of information...yet this is what the cycle depicts. We all know that we are supposed to look at each step and reassess as more clarity is provided. However:

a) this is NOT how the process is taught or practiced
b) we rarely change the collection plan once it is built (or actually ever refer back to it)

Having a 'process' is what is used to train the novice and it is what the novice performs in the field. As the novice grows, nothing is done to change their approach unless they are lucky enough to meet up with a mentor who HAS gone beyond the limitations of school house.

And school house is where the change needs to be that is where the monoliths have their base.

Anonymous said...

Thank you for a great email! It tells all!

An opinion...
The table that in your email that gives a tabular summary of the intelligence functions as a function of the agency tells it all. That single chart says that there is no logic, i.e., systems design philosophy, that guides the structure of a system; therefore, the intelligence cycle is broken or non-existent. Based on that chart, there is no process or procedure that can be defined and defended. "Things" exist that make it work, but in all probably that "system of things" is heavily dependent on expert human intervention. If one or more experts retire or is missing, then the "system of things" breaks down.

I think that single chart gives you carte blanche to system design from scratch a generic Logic based intelligence process and cycle.

Thank you.
Very respectfully,

The chart finally explains why the Advanced Combat Systems community has been unable to communicate with the Intelligence Community. The differences between the two in Science, Engineering, and Technology are like night and day!