Talk:Intelligence cycle management

From Wikipedia, the free encyclopedia

MILHIST This article is within the scope of the Military history WikiProject. If you would like to participate, please visit the project page, where you can join the project and see lists of open tasks and regional and topical task forces. To use this banner, please see the full instructions.
Start This article has been rated as Start-Class on the quality scale.

This is the top of a hierarchy of articles that I'm rolling out, as a response for more high-level discussions as well as the technical details of SIGINT, MASINT, and their subdisciplines. I tried not to be US-centric, but could only find US published sources for a good deal. Material about other countries' programs is more than welcome.

I recognize that even though I'm taking parts out of articles, they are on the long side. Suggestions on splitting also are welcome. Howard C. Berkowitz 01:05, 30 October 2007 (UTC)


Contents

[edit] Which article would "preventing "circular intelligence" or "the daisy chain"" go under?

Not sure if this is covered in other articles - In "Military Intelligence Blunders", by Col. John Hughes-Wilson, he describes how unconfirmed information can inadvertedly become confirmed. Ie.:

1: Agency A reports an unconfirmed fact or assessment.
2: Agency B picks this fact or assessment up, and repeats it as a straightforward report.
3: Agency A then sees the fact in Agency B's report, and seizes on it as independent proof that their information was confirmed by another source.
As a result, ""B-6" information (i.e.: "from a normally reliable source, but not confirmed") suddenly becomes "B-1" information ("from a normally reliable source, now confirmed by other sources")
The book states that this unfortunately happened just before the Yom Kippur War.
1: The Israelis judged (but did not confirm) that war was not likely.
2: The CIA reduced their threat assessment (i.e.: of the likelihood of war) because the Israelis did not seem concerned by the Arab build up. The American's thought that if the Israelis (who had the most to lose from a war, and the best intelligence sources) were not alarmed, then they could downgrade their threat assessments.
3: When the Israelis saw the US downgrade their threat assessment, it confirmed their earlier judgements - if Washington wasn't worried, why should they be?
4: Neither side realised that the other had only downgraded their threat assessments because they thought the other wasn't worried about war breaking out. --Athcnv (talk) 00:40, 19 November 2007 (UTC)
Good points about a very real problem. I believe this is most likely to be caught, at least in terms of the hierarchy of articles on intelligence, in two places, depending on the timing of the two reports.
  • intelligence collection management. In your example, it is the responsibility of the people that do the collection department markings to be sure that "confirmation" is truly that. In your example, when Agency A is holding the report at B-6, when it wants to confirm it with the Agency B report, Agency A is responsible for backtracking the Agency B report until the actual source is known. If, as in this case, the report is from the estimate of a second country, or the other country will not reveal its source, I would say that the B-6 remains B-6, although there can be a footnote citing that another service believes, for unknown reasons, that the report is true. The report, with footnote, then goes to the analysts.
  • intelligence analysis management. Let's assume that the CIA Operations Center, as it probably does, maintains situational awareness of the threat of war in the Middle East. In other words, there is a "threat level", but not a "report". If a policymaker asked the Operations Center about the chance of war, it would be reported as a warning condition -- not a report reliability code.
Now assume that there is an analyst that receives a status report from the Station Chief in Tel Aviv, with an assessment that war is not likely. That analyst is responsible for putting things together before passing it on, assuming there is an information exchange with Israel. I would regard it as irresponsible for the analyst to pass an unconfirmed report to the Israelis. As I mentioned before, if the confirming information cannot be identified, it is probably not prudent to pass along a confirmation from an unknown source.
If, by any chance, you are familiar with Internet routing, there's a design issue in certain routing protocols, which is rather like this. Some routing protocols (i.e., the link state class such as OSPF and ISIS) recalculate the best route to a definition from completely raw data each time there is a computation. Some other protocols (i.e., the early distance vector class such as RIP), incrementally add their costs to get somewhere, assuming the information they are given is accurate -- but it may be badly out of date.
There are a number of ways to avoid this problem, one of which is called "split horizon". Simply stated, if you learn about a destination from one source, you never tell that source any more about that destination. The architect assumes that the source closer to the destination has a better idea what is going on, so the bureaucrats up the food chain have no business contradicting it.
Did the split horizon example clarify or utterly befuddle the point? This would be a lot easier with a drawing. Not implementing split horizon, without other safeguards, tends to cause the "roadmap" of the network to turn into an infinite loop. Howard C. Berkowitz (talk) 00:58, 19 November 2007 (UTC)
I thought of a way to restate it. Once you have sent an estimate to someone, you don't consider estimates returned by them unless they identify them as being based on source information that you don't have. It sounds like what you were describing were estimates being changed in response to other estimates. Howard C. Berkowitz (talk) 21:55, 21 November 2007 (UTC)

[edit] Wording of introduction

This is in response to:

  • 04:10, February 26, 2008 Hcberkowitz (Talk | contribs) (63,724 bytes) (I'm trying not to get into an edit war here; please discuss concerns. Minor: "In one documented case" makes "for example" redundant -- and there were multiple such cases)
1) This is a good article overall.
2) The introduction, though, was very poorly phrased. I think I have improved that phrasing with my edits so far.
3) Yes, the specific "undo" you did is in this case very minor and not worth an "edit war." Still, I stand by my original wording. "In one documented case" does itself serve as a sort of "narrative bridge" from the general to the more specific, it's true, but the "for example" is still necessary, stylistically, to reinforce that transition and to relate this sentence to what has preceded it (ie, that this is a specific example of how intelligence can serve the national interest). Otherwise, without both phrases, the transition is jarring and abrupt. There should not be any reason to disagree, especially as it is, again, a very minor point.
4) I think the phrase "(and sensors managed by intelligence specialists)" is redundant an unnecessary and would recommend its deletion.
5) What were the other "multiple such cases" that resulted from my attempts to improve the writing in question? I'll be happy to clarify any other issues you may have...
--Wikiscient (talk) 04:33, 26 February 2008 (UTC)
Let me explain some of my concerns. I have yet to hear what you found wrong with the introduction other than "poorly phrased". I don't agree with you about the "In one documented case" matter, but it is too small to worry about. I stand by my original wording, which did not get into the nuances of whether it was one case or more cases, which I don't think is necessary in an introduction to a much more general article. I do not believe either the "in one documented case" or "for example" improves the original wording, but I'm not a member of the grammar police.
My greater concern was alleviated when you did not undo my changes regarding the perversion of intelligence, which is, I believe, a key image that helps set the mindset for the closely following section on failures. There is, incidentally, an extensive literature on intelligence failures, with some brief sourcing in this introductory article, but other failures being the topic of entire general articles (e.g., cognitive traps in intelligence analysis), playing a major part in other articles (e.g., Battle of Pearl Harbor), and, of course, books such as Wohlstetter's.
Also, I appreciate not reverting the Sun Tzu and Richelson examples. I was not sure why you originally commented them out; I think they are useful examples.
In the case of near-accident nuclear war, incidents have been reported by both the US and fUSSR. From memory, around 6-8 incidents have been described. The case where BMEWS announced what appeared to be a large Soviet attack, but was in fact a phenomenon resulting from the radar not having been tested against a particular phase of the moon, caused an immediate war alert within the NORAD Combat Operations Center. The on-duty watch officer, however, knew that Khruschev was at the UN, and, from national intelligence sources, had no reason to believe the Soviets had sufficient internal turmoil to attack the US and sacrifice him. By the time the BMEWS data became seriously anomalous, predicting more launches than the Soviets conceivably had (that itself being STINFO and order of battle intellgence), and refusing to predict impact points, that would have been uncomfortably close to when alert bombers should have been in minimum interval takeoff, and, IIRC, awfully close to when ICBMs should be launched.
The officers that made the holdback decisions sometimes held positions that blurred operations and intelligence. That sort of blurring, for example, is why I made the qualification of sensors that may or may not be operated by intelligence personnel, such as early warning radars under NORAD control. Some sensors are considered organic to operational organizations. There can also be a blurring, for example, between organizations that do not necessarily think of tactical scouting units as intelligence, reserving the latter term for analysts and staff. The confusion over such awkward acronyms as C4ISTAR illustrates the blurring.
The subject of near-accidental, and the hypothetical accidental, nuclear war does not need to be detailed in the introduction to a series of articles; there could well be an article on the declassified cases available, and there are several books and monographs on the subject.

Howard C. Berkowitz (talk) 04:52, 26 February 2008 (UTC)

-I have yet to hear what you found wrong with the introduction other than "poorly phrased".
That's all at this point. The content itself, as it stands now, is ok.
-I'm not a member of the grammar police.
I am. I also care about spelling, style, clarity, relevance, and all the other issues addressed at WP:MoS.
Let me rephrase that as "style police".
-In the case of near-accident nuclear war, incidents have been reported by both the US and fUSSR. From memory, around 6-8 incidents have been described.
I assumed the citation was for only one of those. (The one that stands out in my own mind -- and I don't recall the source at the moment, I think a New Yorker article -- was when a Soviet duty-officer watching over a relatively new satellite-based early-warning system got multiple launch alarms and took the personal responsibility of not sending it up the chain of command until after it would have been too late to respond anyway -- it turned out to have been sunlight reflected off of clouds, and they had to recalibrate the system to deal with that -- I forget what happened to the duty-officer, but he most likely ended up spending some time in Siberia after that!)
-the Sun Tzu and Richelson examples
Those are fine, now that they're referenced. They could be better incorporated into the rest of the introduction, probably, but they're fine as they are.
A little confused -- they were sourced before.
-nuclear war does not need to be detailed in the introduction
I would not complain if the whole "example" were deleted.
I would, and let me explain the reasoning, which, to some extent, was discussed among several people when creating the series on intelligence. Prior to starting on what was called an "overarching" article to give structure (I regret I don't remember who came up with the term, which wasn't mine. Might have been ajr or AzureCitizen), there were, and still are a few slithering around, lots of articles that seemed primarily sourced on James Bond.
If I were writing this article for a class, or perhaps in an intelligence journal where I would expect a longer attention span, I would agree completely with you. Wikipedia has an assortment of conflicts, one of which being a belief that a nonspecialist can get a general idea of why a subject is important, and the basics of how it works. Of course, this doesn't get very far when the article involves a subject that can't even be discussed without a good deal of mathematical or scientific background. I have taken a break from editing network engineering and computer science, because I got very tired of people insisting on oversimplifications that were outright wrong.
Nevertheless, there seems to be a delicate balance point, especially when there is an introductory article, between being precise, and being attention-grabbing to a nonspecialist. Sometimes, it seems to work to have just the lead be somewhat more journalistic that would be ideal in a classic encylopedia. I think you'll find that lead is different than the style I tend to use in the body, or in a more in-depth article. That, in any case, is the reason I wrote it the way I did, and I offer it as something I found I had to learn through experience.
If you think diplomacy may be needed here, you'll just love something like a CIA-specific article and its politics, even if the matter in question might have done by Army Intelligence before the CIA was even formed. :-(
-There is ... an extensive literature on intelligence failures ... other failures being the topic of entire general articles
Yes, and I'm not sure this is the best place to get into such detail about it here. Certainly, that section of the article ought to come after (and as a sub-section of) what this article is actually about -- ie. intelligence tasking and management. I mean, it just seems sort of awkward to me to talk about how that fails before talking about what it is. I'd move the failure section to the end, or even better put it in its own article within the hierarchy-of-intelligence-articles and just reference it here...
Again, consider the general reader. Logically, you are correct. In practice, the consequences of error, I believe, need to be in the introduction to keep the general reader reading.
Despite my user name, I am still a wikipedia-newbie -- I admit it. But I do have a lot of related experience. And this subject does interest me, so I may well join the project itself. It certainly seems to be something that you have yourself been significantly contributing to, so I hope we will not have to be working at cross-purposes over such minor points in the future. I'm just trying to help out for the greater good -- I don't mean to be stepping on your toes in the process!!
Peace,
--Wikiscient (talk) 05:37, 26 February 2008 (UTC)
A good evening (or as appropriate to your time zone) Howard C. Berkowitz (talk) 05:51, 26 February 2008 (UTC)

[edit] But what is it?

The article fails to clearly explain what "intelligence cycle management" actually is. Per Wikipedia:Layout#First_paragraph, the first sentence should clearly introduce the subject. Superm401 - Talk 05:21, 12 March 2008 (UTC)

In my judgment, the introductor material is necessary, and I do not regard a style suggestion as a rule, as Jimbo Wales puts it, that can never be broken. If you have specific suggestions, I will be happy to listen, but I am not going to struggle to rewrite something that many people have found useful Howard C. Berkowitz (talk) 05:56, 12 March 2008 (UTC)
You could add at the very top something like: "ICM is a term describing the top-level coordination of intelligence tasking, collection, analysis and dissemination activities, which are collectively described as the IC", and then continue with the current text. Sv1xv (talk) 05:58, 14 March 2008 (UTC)

[edit] 100% Biased Against Intelligence

This article wreaks of bias against intelligence. It states only numerous failures. At the very least, it needs to acknowledge that the many successes of intelligence activities by local, state, national, coalition, or international intelligence organizations are not able to be listed due to either security reasons or, more commonly, the fact that nothing happened because good intelligence information got to the decision-makers and/or folks in the field who acted accordingly and prevented any mishap from occurring. Come on, let's be fair here. T geier (talk) 11:15, 27 March 2008 (UTC)

I thought it did make such an acknowledgement, and if you feel it didn't, I welcome you to put it in. I wrote much of this article, and in no way do I think intelligence is other than an absolutely necessary government function.
When I put in the information about failures, it was to show a problem in management. If you look, for example, at the sub-articles on SIGINT, especially SIGINT in modern history, there are many recognitions of success. At a minimum, these include the Battle of Midway and VENONA. I am perfectly willing to get examples from other countries, such as the Zimmermann Telegram. IMINT had a very major role, of course, in the Cuban missile crisis, and also in arms control (see National means of technical verification, which I also originated and regards intelligence as a necessary function). There was very substantial analytic work on the Soviet economy, although that's harder to source.
One way of citing successes, without making the article too large beyond an equally general discusion of failures, would be wikilinking to the successes in the sub-articles.
So, first, I don't agree there is 100% bias. Second, what I wrote emphasized the process over success and failure, which isn't always clear, especially in long-term strategic analysis. Third, will you help in putting in examples of successes, under the appropriate topics (e.g., collection successes, analytic successes, etc.)? There are mixtures of success and failure that perhaps should be noted, such as the war warning to Pearl Harbor (successful tasking, collection, and analysis) that did not reach the responsible commanders through a failure of dissemination?
Howard C. Berkowitz (talk) 15:37, 27 March 2008 (UTC)
I will re-read the article, but bias was the first impression I got of it. And the failures resounded more initially than anything else. Sorry if my comment was a bit rash. I will re-read and I would heartily agree to working together on this. Thank you for your execellent response. 07:54, 28 March 2008 (UTC) —Preceding unsigned comment added by T geier (talk • contribs)

[edit] Context and jargon

Hcberkowitz has created a good article bereft of the uncited, unverifiable, US-centric, or biased content typically found in such specialized topics. However, the flow of the article is disjointed and reads more like a professional development text or manual rather than an encyclopedia article for generalists. I have tagged it with the context and jargon templates until some of these issues for a "top level" article can be resolved. Basically, the article is declarative but not descriptive. Historical examples are given within a framework, but the rationale for the framework is justified with lots of jargon. Organizational acronyms and flowcharts are introduced, but they distract from the issue at hand by requiring readers to browse elsewhere to understand an acronym or decipher bureaucratese. Quotes from the literature are interspersed in the text. At the end, I feel I have an incomplete picture of how intelligence is managed given its cyclical nature -- which is what one would expect to learn about given the title. Given the quantity and quality of the information present, the article has definite promise, and I look forward to revising it. Madcoverboy (talk) 01:54, 21 May 2008 (UTC)