The David Kilcullen Collection (merged thread)

[UrsaMaior]

As of soldiers in non-combat duties the aqueduct of Caesarea should not be too far from you Wilf. Some "graffitis" of the legionaries who have bult it still can be seen. The legionares were conducting highway patrols, serving as scriptors in the province's adminstration etc. Dont wanna show off with my weak latin so I just repeat the old roman saying: there is nothing new under the sun.

With respect given where respect is due I think contemporary western privates can do much more.

[Cavguy]

Point being, why are we all trying to define what "Insurgency" is? Is it to somehow validate the current ideas about COIN?



And Niel... I fully accept that you are locked into a organisation with it's own language, and in order to make headway you have to speak that language. Good job and good luck. I merely suffer no such constraint.

Wilf -

Unfortunately I can't blame the Army for this. This is pure academia. One of the basic requirements of my thesis was to define what I was talking about, and how it was (slightly) different from standard civil war studies, which have a defined definition in the PolSci/IR world. There have also been studies of terrorism, which again, are different than insurgencies. So the above paragraph was my attempt to try and clarify what was being discussed. Like you, I'm less interested from a military standpoint in the distinctions.

Rex,

I just checked Lyall's website, he has finally posted the dataset almost a year later. Our protests to IO and USMA must have worked. However, the dataset itself is in some sort of stats file, so I have to figure out the program it's from.

http://www.princeton.edu/~jlyall/Lyall_Newtest_Data.htm

Niel

[slapout9]

To my mind, one of the most telling things (as I have said before) is this is "Small Wars" which is very old language. We might try speaking the old language more and gaining more clarity from it.

Hmmm a cognitive COG by doing that one thing alone we could solve a whole bunch of problems. From the SWJ Homwe Page!

Small wars are operations undertaken under executive authority, wherein military force is combined with diplomatic pressure in the internal or external affairs of another state whose government is unstable, inadequate, or unsatisfactory for the preservation of life and of such interests as are determined by the foreign policy of our Nation.

--Small Wars Manual, 1940

[Mark O'Neill]

On another note, I have a lot of time for Niel's point. As a PhD candidate , definitions and terms are very , very (should I say very again?) important to framing your debate. In all honesty, I after a few years on this site, I can say that it is my belief that that it is no dfferent here.


Bottom line guys - if you want effective arguments, define your terms, clearly and parsiminously,

Cheers,

Mark

[William F. Owen]

With respect given where respect is due I think contemporary western privates can do much more.

Really? I'm not a 100% sure but my guess is that most Legionnaires of the time probably spoke 2 or more languages, and probably had a wide range of other non-military skills. - but to your point, can we do better than the Romans? - yes I'd like to think so, but as concerns our capacity to learn and reason, we are the same people.

... and yes, I go to Caesarea regularly, for the beach, not the archeology!

[William F. Owen]

Unfortunately I can't blame the Army for this. This is pure academia. One of the basic requirements of my thesis was to define what I was talking about, and how it was (slightly) different from standard civil war studies, which have a defined definition in the PolSci/IR world. There have also been studies of terrorism, which again, are different than insurgencies. So the above paragraph was my attempt to try and clarify what was being discussed. Like you, I'm less interested from a military standpoint in the distinctions.

OK, understood, but I would still be inclined to challenge the validity and usefulness of defining something called "insurgency" in a way that is supposed to add value and understanding - because it may do the opposite.

...but your examining professor may well think otherwise!

[Rex Brynen]

Really? I'm not a 100% sure but my guess is that most Legionnaires of the time probably spoke 2 or more languages, and probably had a wide range of other non-military skills. - but to your point, can we do better than the Romans?

OK, I've tried very hard, but I can't resist.

So tell me, Wilf.. what have the Romans ever done for us?

[William F. Owen]

OK, I've tried very hard, but I can't resist.

So tell me, Wilf.. what have the Romans ever done for us?

So remind me. Which Romans are we talking about here ?

[Abu Suleyman]

can we do better than the Romans? - yes I'd like to think so, but as concerns our capacity to learn and reason, we are the same people.

I would argue that not only can we do better than the Romans, but that we do. Even privates whom we consider uneducated, and barely speak English probably have a larger vocabulary than bilingual or polyglottal Romans.

The simplest explanation is that the world is a lot more complicated today than it was two-thousand years ago. After all, the Romans believed that there were only four elements. I am pretty sure memorizing the periodic table was easy. To put it into perspective, the great Roman writers probably knew most if not all of the words in their language, and Classic Latin had a vocabulary in the tens of thousands. Contemporary English has a vocabulary of over a million words. Functionally, every word in classic latin has entered english in some form or another, and we topped it off with Anglo-Saxon, NOrman French, a fair amount of Greek, and any other word we wanted to pirate from other language. Talk about perestroika!

Moreover, if a Roman were to learn a foreign language it would probably be something like other Latin languages, or perhaps Celtic, or Greek which were themselves very simple at the time and were all Indo-European, and less difficult to learn than modern languages from divergent language families. It wasn't until the Romans met with the Arabs, Jews, and Carthaginians that they encountered non-Indo-European languages.

So I would say that we need to cut our soldiers some slack. They may not know as large a percentage of the available knowledge as the average legionary, but at least they know that your brain is not for cooling you blood, and the best way to treat a fever is not to slit your wrists.

[UrsaMaior]

Yeah I was always wondering what an aqueduct was doing on the beach (I know 2,000 years is a lot of time). ;-)

IMHO before good deeds and well in advance of killing there has to be understanding. Understanding of the local's Maslow pyramid. We are way ahead of the romans in psychology.

[JasonLyall]

Dear CavGuy,

Thanks for your continued interest in our work. We're looking forward to reading your reply in International Organization.

For those who are interested, I'd simply note that we posted the replication data, including the exact codes for statistical replication, on my website in April (the article was published in January 2009). In January, we also posted the codebook, supplemental analyses, and cases (286 of them).
http://www.princeton.edu/~jlyall/

Though IO does not have a replication policy, we hoped that this dataset will be used, and improved upon, by other scholars, and so we released it. By making it available to all parties (it's been downloaded more than 50 times since posting), we aim to provide a public good.

Are there limitations with these data? Of course. But rather than flaming us anonymously on a chat forum, perhaps it would be a good idea to open the dataset, look at what we've done, and then make efforts to improve upon it. That way, we all win, and we move the study of COIN forward.

Anyways, just wanted to set the record straight.

[marct]

Hi Jason,

For those who are interested, I'd simply note that we posted the replication data, including the exact codes for statistical replication, on my website in April (the article was published in January 2009). In January, we also posted the codebook, supplemental analyses, and cases (286 of them).
http://www.princeton.edu/~jlyall/

Thanks for making the data sets and code books available. I do have a couple of questions:

1. Why are you starting "RAILROAD" in 1871? Railroads used during the US Civil War, and threatened during the first Riel Rebellion. When you state "The first observed use of railways occurred during France’s 1871 suppression of restive Kabylie in Algeria" (p 1-2 Codebook), this is simply incorrect.

2. In your MECH variable, you "counted the number of main battle tanks, medium battle tanks (1917-45 only), armored personnel carriers (APCs), armored fighting vehicles (AFVs), scout cars, and self-propelled artillery in each country’s arsenal." (fn 2 Codebook; emphasis added) Why are you counting the arsenal value rather than the deployed value? Also, while I can certainly understand the data limitations on trucks, couldn't you have developed a proxy variable based on, say, production / import of trucks?

3. I notice that you are using HELI to refer solely to helicopters. No problems there, but why not a variable for other airpower? I'll note, by way of example, the Nuer rebellion was smashed by highly limited airpower strafing the herds and was the decisive variable in that conflict.

4. In your list of conflicts, why is the Second Riel Rebellion excluded?

Cheers,

Marc

[Steve Blair]

Jason,

I just wanted to toss in a bit along the lines of Marc's comments.

Concur on his point 1. Railroads were used heavily during the Civil War and also factored into Frontier troop movements during the various Indian conflicts. Ref point 3, air support also played a role during the Marine involvements in Central America, bot in terms of what we'd consider attack aviation and supply functions (mainly air drops to isolated patrols, but it was still a factor and worked into operational planning).

I could also quibble about the "Second Apache War," and the timeframe listed for the Navajo War is also incorrect. The main campaigns were roughly 1862-1865. And speaking of the Marines, why are Haiti and the Dominican Republic listed but not Nicaragua?

[Cavguy]

Dear CavGuy,

Thanks for your continued interest in our work. We're looking forward to reading your reply in International Organization.

For those who are interested, I'd simply note that we posted the replication data, including the exact codes for statistical replication, on my website in April (the article was published in January 2009). In January, we also posted the codebook, supplemental analyses, and cases (286 of them).
http://www.princeton.edu/~jlyall/

Though IO does not have a replication policy, we hoped that this dataset will be used, and improved upon, by other scholars, and so we released it. By making it available to all parties (it's been downloaded more than 50 times since posting), we aim to provide a public good.

Are there limitations with these data? Of course. But rather than flaming us anonymously on a chat forum, perhaps it would be a good idea to open the dataset, look at what we've done, and then make efforts to improve upon it. That way, we all win, and we move the study of COIN forward.

Anyways, just wanted to set the record straight.

Dr. Lyall,

First, Thanks for posting your dataset. I know (as this discussion above shows) how hard it is to code one. We thought it might be better than the RAND set we were using, that is why myself and several others wrote you shortly after publication asking for it, hoping we could use it in our graduate work at K-State while attending fort Leavenworth. Dr. Jones and Libliki were kind enough last fall to send us theirs from RAND, even given those limitations. In addition to my request, I even had my KSU professor email you directly for your dataset this winter. As noted about five posts above, I did credit you for releasing the data. (read the whole thread)

Also to be fair, versions of your study have been floating since September 2007.

Second, I am not anonymous on this forum. A click in my signature (Who is Cavguy) will tell you exactly who I am, I have been "out" on SWJ for awhile, and all the regulars here know who I am, with 900+ posts since joining on this forum, where I have learned much. Your "Rage" thesis has been extensively discussed here as well, if you follow this link. You're welcome to join our discussion.

Finally, I will send a separate private email with what was submitted to IO for your comment, on or offline.

Regards,
Major Niel Smith
US Army Command and General Staff College Group 13D

[Cavguy]

For those who are interested, I'd simply note that we posted the replication data, including the exact codes for statistical replication, on my website in April (the article was published in January 2009). In January, we also posted the codebook, supplemental analyses, and cases (286 of them).
http://www.princeton.edu/~jlyall/

As recently as 7 July 2009, your dataset was not linked from your website. Dr. Toronto and I looked often for the data, and given the previous responses from the KSU emails this winter, we assumed it was still being withheld/worked on. Am glad you resolved the bugs and posted it.

Also, I don't think my previous criticism was a flame, but I didn't understand why the data couldn't come out earlier with the paper, since its genesis was over two years old. When that happens, one begins to question the dataset. I also think your paper is critically flawed on a logical basis. When you read the response we authored (I sent via email), we don't challenge the dataset or SATA results, only the reasoning, conclusions, and case study.

I am posting our major arguments over at the Mechanization and COIN thread (see post #99). Hope you join in, there is a lot of very insightful debate/discussion there on this subject that adds experiential context to theory.

Niel

Ibid. On 1 May 2009, this appendix was available—as Lyall and Wilson 2009, 92, report it would be—at http://www.princeton.edu/~jlyall/Lya...t_Research.htm. On 6 May 2009, it was not. The appendix contained a replication of Fearon and Laitin’s 2003 study on intra-state violence; robustness checks of the correlations they report in their article; and a list of their 286 cases, with start dates, end dates, and outcomes. As of 7 July 2009, this appendix was still not available at the above site.

[JasonLyall]

Dear Marct, Steve, and CavGuy,

Thanks very much for your replies and questions. CavGuy, I look forward to reading your piece; thanks for sending.

Still not sure why there's confusion re: the data, codebook, and appendix. I checked, and copies of all are still sitting on my Princeton website, with the dates when they were posted clearly marked. I even sent an email around to all the scholars who contacted me (including some of the MA students in your program) informing them that the dataset was released --- I've been working with people at IDA with it since April.

But as long as you've got the data, great. I'd note, too, that it's customary for scholars to release their data once their article has been published, but not before, since we want to make sure we get a publication out of it (and have the data peer-reviewed) before we release it.

As for the other questions:
(1) On the railway variable. I agree with the early start dates --- railroads played a huge role in the US Civil War, no question. This case isn't include in our dataset, however, since it was fought primarily (though not exclusively) along conventional lines. We tried to ensure that the conflicts we included were fought principally along guerrilla warfare lines. Of course, mixed cases exist, but I'd argue that the US Civil War was principally conventional, so it was dropped from the analysis.

(2) The first and second Riel Rebellions were considered but were excluded because they did not met the COW 1000 battle deaths threshold. The same holds for the Sandino Rebellion (1927-33), which I take as the Nicaragua example that was raised (136 Marines died, for example, and around 500-600 rebels). I'll have to recheck the Indian campaign dates again; thanks for pointing that out.

(3) I totally agree that the air power variable could be expanded. In fact, I'm just back from Afghanistan, where I've started a project on the impact of airstrikes on insurgent violence. So, totally agree here, and I think lots of interesting work could be done.

(4) Why not use the actual numbers of deployed mech? This presents three problems: (1) the data are surprisingly sketchy once the war has started, especially for earlier cases; (2) where you take the measurement (at the war's start? middle? end? an average?) is tricky; and (3) most important, the number of MECH deployed is endogenous to the war itself, meaning that your outcome measure and your measure of MECH become correlated, resulting in mistaken inferences given the selection effects present.

Thanks again for the questions and comments.

[marct]

Hi Jason,

(1) On the railway variable. I agree with the early start dates --- railroads played a huge role in the US Civil War, no question. This case isn't include in our dataset, however, since it was fought primarily (though not exclusively) along conventional lines. We tried to ensure that the conflicts we included were fought principally along guerrilla warfare lines. Of course, mixed cases exist, but I'd argue that the US Civil War was principally conventional, so it was dropped from the analysis.

Okay, I can understand that, but if that is so, why would you have included the Maoist revolution since it went conventional too? It strikes me that there is a very shady area between Phase 1 and Phase 3 wars that could loosely be termed as "insurgencies" or "revolts" that needs to be pulled apart.

(2) The first and second Riel Rebellions were considered but were excluded because they did not met the COW 1000 battle deaths threshold. The same holds for the Sandino Rebellion (1927-33), which I take as the Nicaragua example that was raised (136 Marines died, for example, and around 500-600 rebels). I'll have to recheck the Indian campaign dates again; thanks for pointing that out.

Totally agreed on the first Riel Rebellion - it was more a case of political theatre with only a few deaths associated. The second one, while it doesn't meet the 1000 cut off (directly), though, might be worth considering. I've always been leery of using an arbitrary cut off since I think it skews the data towards kinetic insurgencies. So, for example, insurgencies such as Estonia's Singing Revolution obviously wouldn't be included but, IMO, should be. The asymmetry of such conflicts is not on the battlefield per se, but is decisive in the political ends, and that can have some valuable lessons for other, more kinetically focused, insurgencies.

(3) I totally agree that the air power variable could be expanded. In fact, I'm just back from Afghanistan, where I've started a project on the impact of airstrikes on insurgent violence. So, totally agree here, and I think lots of interesting work could be done.

I look forward to reading it when you come out with it. You might think about parsing your airpower variable by function: movement, recon and strike as examples. If I remember correctly (and i could easily be wrong, it's been a while since I looked at it), in the Nuer war, the Brits had only 4 planes in operation and used them initially for recon. It was only at the end of the war when they started using them for strikes on the herds that they became decisive.

(4) Why not use the actual numbers of deployed mech? This presents three problems: (1) the data are surprisingly sketchy once the war has started, especially for earlier cases; (2) where you take the measurement (at the war's start? middle? end? an average?) is tricky; and (3) most important, the number of MECH deployed is endogenous to the war itself, meaning that your outcome measure and your measure of MECH become correlated, resulting in mistaken inferences given the selection effects present.

A good synopsis of the problems.

1. Agreed, but I suspect that much of it could be teased out. More importantly, the arsenal numbers probably do not reflect the same percentages as are actually being used in the field which throws a pretty serious question on the validity of your argument re: mechanization. It might be possible to take exemplar cases from the various time periods and use those to strengthen your argument.

2. Actually, I think this is a false problem: I would use a slope of initial, middle and end to reflect changes in the operations of the campaigns. You could, in effect, create a matrix of inter-related variables with, say, a "High" initial MECH value which drops during the campaign. This would allow you to have a much stronger argument about the relative value / cost of using MECH in various campaigns.

As an added note, I think you would also get a lot of additional information if you functionally split out the uses of MECH during the campaigns in a similar manner to what i was suggesting for airpower.

3. I think I would disagree with you on this point. I would argue that it is pretty strongly correlated at the start of a campaign, albeit modified by both doctrine and units organization/selection, but not necessarily so throughout the campaign. Selection pressures during the campaign would not be reflected in your model, which would result in mistaken inferences about the causal effects of MECH on the outcomes. In effect, the argument line you seem to be following appears to assume that outcomes can be predicted from starting conditions which ignores the actuality of the campaign process - the "friction" and "fog of war" as it were.

It's really the implications of this third point that concern me more than anything else. Once such a mechanistic (no pun intended ) model is accepted, then, I would argue, it only holds true while all of the sides involved accept the basic parameters of that model. But such a situation is what gives rise to variable asymmetries in the first place: a rejection of those parameters (aka "conventions"). I think that this is pretty clearly illustrated in both the primary AO of AQ (i.e. communications) and during the Millenium Challenge wargame when GEN Van Ripper showed how false those parameter assumptions were.

At any rate, that's moving beyond the scope of your initial paper, so I'll leave it there.

Cheers,

Marc

[davidbfpo]

Discovered on an IT security blogsite ( http://www.schneier.com/blog/ ) an Australian TV debate between the COIN expert and a barrister:

Here, in a fascinating discussion with human rights lawyer Julian Burnside at the Melbourne Writers' Festival, he (Kilcullen) talks about the ethics and tactics of contemporary warfare. Julian Burnside is a human rights barrister and refugee advocate. He is the author of numerous publications and books including On Privilege and Watching Brief: Reflections on Human Rights, Law and Justice.

The weblink is: http://www.abc.net.au/unleashed/stories/s2668177.htm and two versions of the talk, a summary and the full, 56 minutes.

There is an audio only link: http://www.abc.net.au/tv/fora/storie...26/2667749.htm

Both websites have comments and this is pithy:

From the Schneier site 'You can't kill your way out of this problem'.

Clearly some of Bob's thinking has reached outside SWC.

There is a previous thread on Kilcullen's thoughts on COIN, but this debate has a different focus, so a new thread; if you want the previous thread it is: http://council.smallwarsjournal.com/...ead.php?t=7224

davidbfpo

[jmm99]

While the discussion touched on some LOAC issues, it never got off the legal ground since Barrister Burnside set out the framework for the discussion and Soldier Kilcullen is no barrister.

The initial hypothetical (by Julian Burnside, who seems a fairly laid-back individual) is interesting. He had UBL on a dialysis machine in the Tora Bora. Do you kill him or capture him ?

I'd add a few figures to the picture besides UBL, to wit: one of his wives and her female entourage, some children and young adults (the older boys being armed), and his ever-present body guards.

What should you do as "Coalition" commander with full power to decide (no higher micromanagement in the picture) ?

Legally, you first have to decide whether that pastoral picture is a Laws of War (military) or Rule of Law (civilian) situation. That is an important issue for discourse because all Coalition partners in Astan are not on the same legal page. As to the video, it would have been nice to have a barrister presenting the US position - which is based on the Laws of War. I'll spare the legal reasoning; but, to the US, we have Hague, 1949 Geneva and the latter's Common Article 3.

As I see it, compliance with US Laws of War depends on the means on hand that can be used. First, posit direct fires (well-aimed at specific targets). Under CA 3, the women and unarmed children are protected. The armed boys and the bodyguards are lawful targets under Hague & CA 3. UBL is actually an issue - is he rendered hors de combat because of the dialysis ? Possibly not - disconnect the tubes and grab the AK. If he were in a diabetic coma, the answer would be protected status.

Now, let's get back to reality at Tora Bora, where that pastoral picture was not present and indirect fires were the only realistic means at hand (based on "MAJ Fury's" book). What would be likely known is an approximate grid location of combatants (maybe including UBL). So, a bomb or bombs make their descent and hit the target - lawful under Hague if military necessity and proportionality apply, even if some civilians are also killed.

Reality would also include the ROE/RUFs then in place, which may or may not sanction the indirect fires COA - or, the direct fires COA, for that matter.

One could also do a legal analysis of the Rule of Law scenario - looking at the problem from a pure law enforcement standpoint.

The bottom line is that Barrister Burnside's first hypothetical and its variations could span the entire hour of the interview and then some. That would have made a great continuing ed segment for me, but very poor audience TV.

The title ("Does the end justify the means?") is a bit confusing when applied to a military context where ends, means and ways have entirely different meanings and usages from the philosophical question of "end justifying means".

[William F. Owen]

Haven't got time to look at the link, but I pretty much despair once people start trying to put "ethical" frame frame works into these discussions. It leads you down a blind alley that utterly fails to account for the realities of operations and imagines that these things are somehow choices or even controllable - they are not.