Yesterday (June 26, 2008), Andrew Exum posted a blog entry over at Small Wars Journal on Information Operations. As part of the post, he laid out a challenge to SWC members.

My challenge for this website’s readers, then, is the following: what do we, as counter-insurgency theorists and practitioners, mean when we use the term “information operations?” Do we use IO as shorthand for psychological operations and message management? Obviously, our definition of IO is different from the definition officially in use. And I suspect our definition has more in common with the way Hizballah or the Taliban thinks about IO than the way in which the lexicographers of the Pentagon think about IO. (That, by the way, is a good thing.) So who out there can propose an alternate definition, and one which we can offer to those in the field in Iraq and Afghanistan? And is “information operations” even an appropriate term for that to which we’re referring?

So far, as of early morning the following day (June 27, 2008), a rather active thread on the question has emerged. Part of the thread was a comment I posted that tried to pull apart the roots of Information Operations (IO) so that an actual, analytically useful, definition could be produced. After working on the post for half an hour, I realized that it would require a lot more thought and detail, hence this post.

Let’s start off by looking at the current US doctrinal definition of IO from JP 3-13 Information Operations. Here (page ix), IO is described as

the integrated employment of electronic warfare (EW), computer network operations (CNO), psychological operations (PSYOPS), military deception MILDEC), and operations security (OPSEC), in concert with specified supporting and related capabilities, to influence, disrupt, corrupt or usurp adversarial human and automated decision making while protecting our own.

Furthermore,

The information environment is the aggregate of individuals, organizations, and systems that collect, process, disseminate, or act on information. The information environment is made up of three interrelated dimensions: physical, informational, and cognitive.

JP 3-13 also defines specific capabilities (page x):

Core capabilities are

IO consists of five core capabilities which are: PSYOP, MILDEC, OPSEC, EW, and CNO. Of the five, PSYOP, OPSEC, and MILDEC have played a major part in military operations for many centuries. In this modern age, they have been joined first by EW and most recently by CNO.

Supporting capabilities

Capabilities supporting IO include information assurance (IA), physical security, physical attack, counterintelligence, and combat camera. These are either directly or indirectly involved in the information environment and contribute to effective IO.

Related capabilities

There are three military functions, public affairs (PA), civil-military operations (CMO), and defense support to public diplomacy, specified as related capabilities for IO. These capabilities make significant contributions to IO and must always be coordinated and integrated with the core and supporting IO capabilities. However, their primary purpose and rules under which they operate must not be compromised by IO. This requires additional care and consideration in the planning and conduct of IO. For this reason, the PA and CMO staffs particularly must work in close coordination with the IO planning staff.

Is it just me, or are other people thinking that this is a smorgasborg of tools, techniques and intentions (TTI - see, academics can create acronyms too!) that are naively tossed together without an overarching theory behind them? And, just in support of this, I’ll note that Andrew used the word defined in his blog post, while the actual word in JP 3-13 is described. Personally, I think there is a major difference between the two (and the OED agrees with me ), and that the choice to use the term “described” is a recognition, at least sub-consciously, of the non-theoretical basis of the doctrine presented in JP 3-13.

Let’s start with a few, common sense, definitions.

Information Operations: operations (actions) that take place within the informational environment

Information: “a difference that makes a difference” (from Gregory Bateson, “Form, Substance, and Difference”, in Steps to an Ecology of Mind, p. 448-466; Amazon, Google).

On the whole, we can use the JP 3-13 definition of Information Environment with a few modifications (i.e. dropping the last sentence):

The information environment is the aggregate of individuals, organizations, and systems that collect, process, disseminate, or act on information.

So, let’s start pulling this apart. If information is a difference that makes a difference, then we have to disaggregate it into two sub-systems: an initial “difference” and as something that “makes a difference”. The first refers to the act of sensing something that is “different” which, in turn, implies expectations about the area of the information environment being sensed. The second refers to an act of categorzation on the part of the actor/sensor as to the “quality” of the difference, and that implies both an analytical schema and an intentionality on the part of the actor.

Let’s suppose, as a highly simplistic example, that I am scanning a bowl of jelly beans. I am sensing many different pieces of “information” about the jelly beans - colour, scent, relative shape, positioning in the bowl, etc. Now, if my intentionality is the eat jelly beans then, according to my own personal analytic schema, there are only two differences that makes a difference (i.e. pieces of information I require to complete the action):

1. Are there any jelly beans in the bowl? (you can’t eat jelly beans unless they are present), and
2. Are there any black jelly beans in the bowl? (personally, I hate black jelly beans and will not eat them if at all possible).

So, first having assured myself that my sense report jelly beans in the bowl, I commit my first action - reaching into the bowl and grabbing a handful. I then do another quick scan of the handful of jelly beans, reassuring myself that there are no black ones, and then pop them in my mouth and eat them.

This example, while silly, is an Information Operation, i.e. an operation where actions are based on assessments of sensory information and an analytic processing of that information based on a schema (a “map” in the sense of Bateson / Korzybski). Of course, I doubt that this operation would be classified as IO by the US military - which would be a mistake on their part.

So, what did the Jelly Bean IO process illustrate?

1. Sensory operations on a given slice of “objective reality”. The act of sensing involved the use of different sensory mechanisms and different channels of sensing (EM waves [visible light], scent, etc.), initial analysis of EM patterns and classification into “colours”, “shapes”, etc., all feeding into an initial “mapping” of the bowl of jelly beans through a process of information classification, recognition and sorting. As an additional note, it is also important to realize that colour is used as a proxy for taste which is a mapping convention (e.g. “black = liquorice”).
2. An intentionality on the part of the actor. In this example, I wanted to eat jelly beans and my internal “jelly bean” map says that all jelly beans, except black ones (a proxy for liquorice), are edible. But what if I had changed that intentionality? What if I wished to eat my jelly beans in order of taste? Then my actions would have been different, and would have involved picking out jelly beans of the same colour and only the same colour to eat at the same time (remember that, with jely beans, colour is a proxy measure of taste).

So we have

1. The act of sensing which includes
   1. using different sensory channels,
   2. preprocessing the sensory information
   3. classifying the sensory information
   4. using proxy conventions to conduct an initial interpretation of the sensory information.
2. application of an intentionality, which involves
   1. selection of a sub-set of potential action
   2. following through on that action
   3. gaining an experience based on that action

Now, this final point of gaining experience is crucial since this is an act of information validation or assurance to use JP 3-13’s language. In reality, it is not an act of information validation but, rather, an act of map validation. Remember that in this example, colour is a proxy for taste, and this proxy is, in actuality, a cultural convention (another possible one could be based on shape). The experience gained validates, or invalidates, this cultural convention and may in fact shift what is “information” by changing the last part of it, i.e. the part of sensory input that “makes a difference”. What if a different convention, say shape, were used rather than colour? If that were the case, colour would now make less of a difference and shape would make more.

Which, in a round about, academic way, brings us back to the military version of Information Operations and what is needed for a really good, theoretically informed, definition of them. Let’s start with a more general definition of Information Operations, and then proceed to a more limited, military version.

Information Operations are a) actions taken by actors, b) based on sensory input from the environment which is c) filtered through one or more interpretive maps, with d) an intentionalty towards a desired outcome.

When the set of intended consequences is to change the actions of other actors, then point d) needs to be modified as follows:

d) an intentionality to modify either the sensory environment, input or interpretive maps of another actor.

Note that this particular modification covers advertising, religious / political / ideological prosletyzing, trying to “change someones mind”, etc. For the more specific, military version of IO, section d) should be modified to

d) an intentionality to either modify, deceive or degrade a targets sensory environment, input or interpretive maps while, at the same time, preserving ones own.

This would give us the following as a definition of military Information Operations:

Information Operations are a) actions taken by actors, b) based on sensory input from the environment which is c) filtered through one or more interpretive maps, with d) an intentionality to either modify, deceive or degrade a targets sensory environment, input or interpretive maps while, at the same time, preserving ones own.

If we take this definition, then all of the different “capabilities” listed in JP 3-13 may be seen as either map, intentionality or sensory system specific Tactics, Techniques and Procedures (TTPs).