Introduction
Combating insurgencies with conventional forces has long been regarded as being, to paraphrase T.E Lawrence’s colorful formulation, comparable to eating soup with a knife (Lawrence, 1922, 53). Indeed, the inutility of force with regards to combating a phenomenon that primarily exists in the minds of a target population has been noted by figures from General Rupert Smith to General David Petraeus, the latter articulating this principle as a central premise upon which he built his population centric theory of counterinsurgency in FM-3-24 (Petraeus, 2006, 60-100) (Smith, 2005, 40). Within the context of this argument, any effort to destroy an insurgent militarily by a policy of attrition or annihilation ignores the insurgents innate capacity to trade space for time, avoiding the strengths of a conventional force and eroding both its domestic will and its control over the target populace (over which the insurgent and counterinsurgent force are fighting) by policies of assassination, intimidation of the counterinsurgency’s local supporters and dispersed attacks on occupying troops. The ability of an insurgency, even one which has held territory for a significant period to revert to what T.X. Hammes dubs phase I of an insurgent strategy (whereby it resorts to asymmetrical warfare) is central to the argument regarding the inutility of an enemy-centric Clausewitzian approach to combating insurgencies (Hammes, 2006, 50).
Rather, it has been contended, counterinsurgent forces must focus on seizing territory, clearing it temporarily, and using the time bought to create the institutional structures that offer long term legitimacy to the counterinsurgency (Petraeus, 2006, 6-10, 77). This approach, sometimes coupled with intimidation of the local populace (an alternative to winning hearts and minds that nonetheless acknowledges the centrality on an insurgencies links to the population) are collectively dubbed indirect counterinsurgency strategies by scholars such as I.A. Toft (Toft, 2005, 12). If a counterinsurgent meets an indirect strategy with a direct one based on either attrition or annihilation, it is doomed to fail while efforts to resolve the inherent asymmetry of the conflict by the adoption of an indirect approach that mirrors that of the insurgent in its emphasis on population control. Within this framework, scholars such as Toft offer falsifiable predictions regarding the prospects for success or failure in three plausible contingencies- one in which a direct strategy by a counterinsurgent confronts an indirect insurgent strategy and one in which both parties adopt an indirect approach. While not disputing the importance of these arguments, I contend that this vein of research, with its implicit assumption regarding the static nature of an insurgent’s strategy, has overlooked a third contingency in which insurgents opt to adopt a direct defense and an attendant empirical puzzle - namely the high occurrence of instances in which the insurgent has adopted this approach, despite its clearly suboptimal nature.
While scholars since David Galula have acknowledged the axiomatic point that when an insurgent, invariably the weaker actor, adopts a strategy of conventional defense the greater power of the state can be brought to bear decisively against it, there has been precious little research into why those insurgents who have adopted these strategies have done so despite the obvious risks (Galula, 1967, 55). Specifically, the question of whether the counterinsurgency’s own strategy can incentivize the adoption of a clearly suboptimal policy by the weaker party in the dyad has not been systematically explored. If indeed a counterinsurgency can shape the way its opponent fights and effect a strategic interaction between two direct strategies, it would follow that perhaps one can indeed “eat soup with a knife”. It is precisely this contention that this article makes.
Building on Daniel Kahneman and Amos Tverskey’s prospect theory, which contends that actors make risk acceptant deviations from rationality when placed in a frame of losses, I will contend that the counterinsurgents actions at the grand strategic and operational levels can achieve precisely this framing effect- placing the leadership of the insurgency in a frame of losses and thereby leveraging the insurgency into fighting on the stronger actors terms. In order to test this thesis, I intend to utilize an in-depth case study of a war which demonstrated the outcome of interest (the adoption of a conventional defense by an outgunned insurgent force). The case study I have chosen is Eelam War IV, which saw the government of Sri Lanka achieve a resounding and rapid victory over the LTTE. While a single case study-based methodology has drawbacks, the study of the process by which an outcome occurred, which involves complex causal mechanisms, requires a depth typically provided by process tracing and absent in more experimental methods (Davidson, 2011, 20). Additionally, given that the LTTE had enjoyed substantial success against the Sri Lankan army by utilizing an indirect strategy in the three preceding Eelam wars, it had every incentive to adhere to an approach that had served it so well- rendering this a least likely case for my argument, insofar as the LTTE had strong demonstrated incentives to adhere to its existing approach and implying that a model applicable to this case will be even more readily applicable in other circumstances. As such, there seems not to treat the Eelam War IV as being sui generis and nongeneralizable. Finally, the findings of a single case-oriented study being used for theory building (the emphasis of this paper) can be treated as diagnostic and used as a basis for further experimental research (Bennett, George, 2005, 13). In the next section, I will lay out my theoretical framework and hypothesis. Following this, I will test the operation of my hypothesis within the context of Eelam war IV using process tracing and the congruence method. Finally, the article will present my quantitative analysis of the correlation between certain types of state action and the adoption of suboptimal military strategies by an insurgency. The final part of the paper will then deal with the possible conclusions that can be derived from this research.
Theory and Hypothesis
The theoretical underpinnings of this article have their origins in Kahnemann and Tverskey’s exposition of risk aversion and acceptance as articulated in prospect theory. Prospect theory postulates that the level of risk that an actor is willing to accept depends on whether the game is framed to him or her as one that entails prospective losses or as one entailing prospective gains. Specifically, if the actor frames the game as one of gains (where he or she is mediating between several positive prospects) the actor tends to be risk averse, whereas a game framed in terms of losses tends to induce risk acceptance. The authors illustrate this through an experiment whereby a set of actors are, in one instance offered a certainty of winning a small amount of money versus the possibility of winning a larger sum, whilst in the other they are offered the certainty of losing a small amount of money versus the probability of losing an even larger sum. What they find, with relative regularity is that actors in the first experiment choose the certainty of winning a small amount (the risk averse option) whilst in the latter they choose the probability of losing a larger sum (the risk acceptant option). Risk acceptant deviations from rationality, then, represent the actions of an actor who is in a frame of prospective losses (Kahneman, Tversky, 1979, 264-273). This may explain, for example, the observed tendency of risk averse actors that can be deterred from conquest to show a willingness to run enormous risks to retain conquests once they have already occurred (Morgan, 2003, 20-22). It is my contention that conceiving of the adoption of a direct defense as a risk acceptant strategy adopted by an actor in a frame of losses offers us a core insight into how an insurgent can be compelled to adopt a conventional defense.
The ancient Chinese strategist Sun Tzu famously stipulated that an attack on an opponent’s strategy rather than his army amounted to the acme of skill in warfare (The Art of War, 20). Interpreting this somewhat cryptic statement, M. Handel argues that Sun Tzu effectively argued that the use of grand strategic tools to alter the strategic environment prior to a conflict allows an actor to shape the psychology of one’s adversary and, by extension, the nature of the conflict (Handel, 2000, 160).
Success for an insurgent depends, as Mao Zedong noted in 1937, on publicly visible victories over the state that effectively problematizes the states authority over a contested area. These victories may amount to small but visible acts (such as the assassination of officials) or large tactical victories (such as the destruction of a body of the state’s conventional forces) (Zedong, 2012 ,68-70). Central to this strategy, however, is the contested nature of control over a specific geographical area. Consequently, my hypothesis is that by establishing a cordon sanitaire over a contested area and withdrawing the states troops beyond this point, the state reframes the contest in the eyes of the insurgency’s leadership shifting them from a frame of gains to one of losses. While counterintuitive, the efficacy of this approach stems from the fact that withdrawing tactically to positions of strength deprives the insurgent of the ability to achieve visible victories over the state in the contested area- an act that underpins its ability to create a sense of building momentum. Secondly, however, by transforming the contested zone into one of insurgent control, it ceases to be framed as a prospective gain and is instead conceived by the insurgent as ground already held (a prospective loss). Crucially, I do not argue for the acknowledgement of insurgent control but rather a tactical withdrawal leaving the insurgent facing the prospect of future government incursions. Secondly, I hypothesize that the establishment of a cordon sanitaire at both the operational level and the grand strategic level (by cutting off external sources of finance and supply) places the insurgent further into a frame of losses, insofar as the insurgent is faced with the prospect of having a diminishing ability to hold its territory as time wears on. Consequently, my first hypothesis is that by a policy of tactical withdrawal and containment at the grand strategic level, the counterinsurgent can steadily shift an insurgency from a frame of gains to one of losses.
Secondly, the contours and cohesion of intra-elite ties within insurgencies are often a function of distributional politics within the group (Staniland, 2010, 55-60). It follows, then, that a policy of containment imposes scarcities and distributional conflicts that exacerbate nascent intra elite friction-allowing the counterinsurgent to leverage away sections of the insurgent’s elite base by economic and political inducements. Simultaneously, the removal of a direct counterinsurgent presence in the insurgencies territory removes any sense of immediate danger that may have lent an insurgency cohesion despite distributional disputes. This further places the insurgent in a frame of losses, less because potentially dissatisfied elites are necessarily vital to the insurgencies functioning, than because the specter of defections reinforces the frame of losses through which the insurgencies core leadership is placed when it is transformed from a challenger to an incumbent. Collectively, I hypothesize that the twin policies of containment and wedge politics at the grand strategic level place an insurgent in a frame of losses that will ensure that when the state chooses to end its policy of containment and assault insurgent held territory, the insurgent will respond with a direct defense (a risk acceptant strategy).
The central prediction made by this argument is that if a counterinsurgent transitions to the offensive after a period of containment, it will be confronted by a risk acceptant insurgency which will face it conventionally. Within this context, contra the predictions of an emphasis on maneuver and firepower will likely to prove highly effective against an insurgency that has effectively evolved into a weak conventional force. Despite the demonstrably poor record of such forces against insurgencies, their efficacy can be radically improved if the insurgent leaderships psychological milieu lends itself to the adoption of suboptimal strategies- an outcome that the counterinsurgent can render more likely by the adoption of a temporary strategy of containment.
Two plausible counterarguments exist regarding this hypothesis. The first is that having assumed the role of a governing body in the territory it occupies, an insurgency cannot readily transition back to guerilla tactics. If indeed this is the case, then the use of prospect theory is superfluous to explaining the adoption of suboptimal strategies by insurgent leaders. This objection might be countered by noting a litany of cases in which insurgent leaders, having seized territory, reverted to Mao’s “Phase I” strategies when they deemed doing so t be efficacious or necessary. Examples of this include the leadership of the fledgling Indonesian republic who, during the Dutch Police Action of 1948, abandoned their then capital of Yogyakarta, along with other cities that they held to revert to the use of guerilla tactics, the Chechen leadership during the first Chechen war and the Taliban in Afghanistan who frequently revert to guerilla warfare after periods of de facto governance over an area (Vickers, 2005, 111).
A second potential objection is that the use of conventional fire power centric military strategies is often, though not always, accompanied by the use of punitive strategies against civilians- and that it is the latter strategy that is central to the brevity of conflicts in which conventional forces do defeat insurgencies in short order. However, the empirical record of the success of civilian targeting is mixed at best, with some studies demonstrating its efficacy while others argue that it is, if anything, counterproductive. Moreover, the plausibility of this potential alternative hypothesis can be reduced by selecting a longitudinal case in which the state’s COIN strategy varies even as its willingness to engage in civilian punishment is invariant.
Having articulated my hypotheses, I will now move on to my case study, where I hope to demonstrate that containment and wedge politics at the grand strategic level, coupled with an operational strategy that allowed the GOSL (Government of Sri Lanka) to operate within its insurgent opponents decision-making loop once hostilities had been resumed allowed the Sri Lankan Government to effectively predetermine LTTE leader Vellupillai Prabhakaran’s fatal decision to adopt a conventional strategy of massing firepower around entrenched positions in a scenario where he was hopelessly outgunned- leading to the rare outcome of an insurgency being militarily defeated.
Case Study-Eelam War IV
The Eelam War IV, which saw the utter rout of the LTTE insurgents who had held the Sri Lankan government at bay for three decades was remarkable for the fact that it seemingly obviated long held nostrums regarding the invincibility of insurgencies in the face of kinetic conventional operations. Indeed, shortly prior to the outbreak of hostilities, analysts such as N. Manoharan of India’s Centre for Land Warfare Studies were offering dire prognoses of the GOSL’s chances of victory in any assault on LTTE held territories in the north and east of the island (Manoharan, 2007, 1). Yet the conflict saw the what has been dubbed the world’s premier insurgency, which maintained networks spanning up to 44 countries and had the distinction of being the sole insurgency able to wage asymmetrical warfare both at land and at sea (via its naval wing the sea tigers) effectively routed by what A. S. Khan dubs an “enemy centric COIN strategy (Khan, 2015, 128) (Roseneau, 2007, 12). It is my contention that the GOSL’s victory can effectively be explained through the application of prospect theory. Specifically, I offer three hypotheses
The acceptance of a ceasefire by the GOSL in 2002, which allowed it to withdraw its forces to areas firmly under its control deprived the LTTE of a foe within striking distance while forcing it to view the areas in which it operated as territories held rather than contested-changing the way Prabhakaran framed the scenario.
The GOSL’s efforts to contain the LTTE, centered around garnering international support in cracking down on the LTTE’s overseas financial networks, building a navy capable of interdicting the LTTE’s overseas arms supply and establishing positions along the juncture between the north and east of the island had the effects both of placing Prabhakaran further in a frame of losses and making the establishment of authority and redistribution of resources to the east of the island difficult- encouraging factionalism in this part of the island and the eventual defection of Prabhakarans subordinate Colonel Karuna-which in turn further entrenched Prabhakaran in a frame of losses
That once the final assault on the LTTE began, the operational decision to carry out simultaneous attacks along a broad front, coupled with the tactical use of 8-12-man SIOT squads allowed the Sri Lankan army to operate within Prabhakaran’s decision-making loop-precluding a reversal of course after the LTTE’s initial failures.
Before delving into my theory, however, I will briefly contend with the two-alternative hypothesis that might problematize my argument. The first argument centers around what has been dubbed the Manwaring paradigm. This paradigm, articulated by the military theorist Max Manwaring contends that the success of a counterinsurgency effort rests on six factors that determine whether the counterinsurgent will substantively weaken the insurgency (Manwaring, Joes, 2000, 100-110). The factors are
Legitimacy
Unity of effort
Discipline and capability of the military
Type and consistency of foreign support for the counterinsurgent
The counterinsurgents ability to cut support to the insurgency
The availability of reliable intelligence
My first qualm with this paradigm and similar approaches is that, as W.C. Fuller notes, they conflate the underpinnings of a strategy with the strategy itself. For example, building competent forces and ensuring the availability of intelligence are certainly axiomatic prerequisites to the execution of a strategy, but they cannot be conflated with the strategy itself (which, pace Clausewitz, entails the direction of these means) (Fuller, 1998, 228) (On War II, 90-99). There is one seeming area of overlap between my hypothesis and Manwaring’s, however- namely the emphasis placed on securing external support to the host government and depriving the insurgency of its sources of external support. However, a crucial distinction between my theory and Manwaring’s lies in our respective assessments of the impact of these actions. As per Manwaring’s thesis, these actions substantively weaken the insurgency to a point where its capacity to resist is seriously debilitated. By contrast, my contention is that these actions do not deprive an insurgency of the capacity to fight a protracted war in the near to medium term, but rather alter its conception of its long-term prospects and by extension its selection of strategy- a fundamentally psychological effect. A cursory examination of the Sri Lankan insurgency does seem to lend credence to the argument that the LTTE was not left substantively bereft of means following the Sri Lankan governments efforts to rally international support to its cause. For example, it is noteworthy that much of the external support lent to the GOSL by actors such as India, the United States and Pakistan in terms of military aid (either covertly or otherwise) was broadly available during the third Eelam war waged at the turn of the millennium- a war which had drastically different outcomes (Mathur, 2013, 128). Admittedly, obtaining the support of actors such as the U.S and E.U in shutting down LTTE financial networks was a coup that the GOSL only achieved in the interregnum between the two wars. Crucially, however, B.C. Tan and J. Solomon estimate that roughly 70 percent of the LTTE’s arms supply consisted of weapons expropriated locally from the Sri Lankan army, while sections of the LTTE’s war effort (such as its seagoing craft) were being produced within LTTE held areas of Sri Lanka (Tan Soloman, 2007, 12). Furthermore, a 2006 report by human rights watch highlighted the fact that the LTTE was at the time able to solicit or extract finances from the Tamil diaspora on a significant scale within the E.U despite ongoing sanctions (Human Rights Watch, 2006). While it seems plausible to assume that the cumulative effect of Sri Lanka’s various efforts at containment might slowly emaciate the LTTE’s financial capabilities over the long term, it does not appear, pace Manwaring, that the organization was precluded from fighting a guerilla war in the near term.
A second hypothesis, raised by figures such as N. Bizouras, postulates that by Eelam war IV, the brutality of the LTTE (particularly evident in practices such as the impressment of child soldiers) had lost it the legitimacy that it had once enjoyed amongst its Tamil constituents, rendering it an insurgent force alienated from its constituency –veritable suicide as Mao would have noted (Bizouras, 2014, 80). While it is inherently difficult to measure the level of popularity enjoyed by an insurgency accurately, it is worth noting that many of the practices that Bizouras associates with the LTTE’s loss of legitimacy were very much a part of its policies during the 1990’s- the halcyon period of the LTTE’s legitimacy as Bizouras would have it. Both the impressment of children for the infamous “baby brigades” and the use of child suicide bombers were an integral part of Eelam war III. Indeed, the LTTE’s signal victory of the war, Operation Ceaseless Waves, was achieved in no small part due to the tactical use of underage suicide bombers (Singer, 2006, 80-90). It is unclear, then, that the LTTE became significantly more brutal in the interregnum, to the point that it alienated constituents who seemed willing to tolerate its brutal rule in the 1990’s.
Having discounted the alternative available hypotheses, then, I will now offer an interpretation of the conflict through the lens of prospect theory.
At the grand strategic level, the debacle of the third Eelam War seems to have forced the GOSL into inadvertently adopting an optimal counterinsurgency strategy. Specifically, the desperate condition of the Sri Lankan military and army (the latter had been humiliatingly routed at Mullaitivu) led the Sri Lankan government to adopt a 2002 ceasefire agreement that effectively saw GOSL forces abstain from entry to de facto LTTE territory (Castellano, 2014, 131). Crucially, however, this decision although not strategic in and of itself, allowed the GOSL to abstain from providing Prabhakaran and the LTTE with a conflict that reified the organizations internal cohesion, whilst giving the SLAF (Sri Lankan Armed Forces) and the GOSL time to articulate a dual policy of containment and reform. At the grand strategic level, the GOSL was able to take advantage of the advent of the war on terror to more aggressively lobby for foreign (particularly western) countries to crack down on LTTE funding. While a number of foreign governments had proscribed the LTTE by the turn of the millennium, the organization had still retained the capacity to garner funding under the auspices of front organizations posing as, amongst other things, NGO’s. However, in the wake of Sri Lanka’s more active diplomatic policy, a number of nations made substantive efforts to crack down on LTTE funding. Notable among these efforts was Canada’s 2005 decision to close a number of NGO’s affiliated with the LTTE and the E.U’s 2006 decision to proscribe the group as a terrorist organization- a decision that precipitated a slew of investigations of LTTE related assets in nations such as France, where the head of the Tamil Rehabilitation Organization was arrested with 18 million dollars of illicit money (Clarke, 2012, 45-50) (Ridley, 2012, 28). Similarly, within Asia, nations such as Thailand followed suit in proscribing the substantial LTTE affiliate organizations on their soil in 2003, while India agreed to heighten naval supervision of illicit supplies leaving its shores from the coastal state of Tamil Nadu (Bhonsle, 2011, 108). Finally, the Sri Lankan navy was subjected to a tactical and operational overhaul to enable it to interdict the LTTE’s sea lines of communication and supply. At the operational level, the newly appointed commander of the Sri Lankan navy Vice Admiral Karranagoda shifted the emphasis of the navy from interdicting smaller ships carrying arms to the LTTE to a CONOPS centered around tackling the large “floating warehouses” from which these ships received their cargo- a shift that required the navy to operate considerably further from its shores than had been previously countenanced. Secondly, the navy adopted the “small boats concept” at the tactical level-substituting large numbers of small indigenously made boats for the high value Dvora craft that had proven especially vulnerable to LTTE swarming tactics. The cumulative effect of these reforms was to further enhance the GOSL’s ability to target the LTTE’s resource base (Mehta, 2010, 128).
Crucially, as has been discussed above, none of this substantially compromised the LTTE’s capacity to govern its territory. Rather, the absence of a military target, coupled with a slow but nonetheless perceptible tightening of a broad financial and diplomatic noose had the effect of problematizing Prabhakarans long term prospects of either moving forward towards de jure independence or retaining his current hold over the north and east of the island. That the LTTE suffered from no material deficit per se is evidenced by its ability to field a force comparable to the one it had fielded in Eelam War III, combined with the discovery of large amounts of previously stockpiled armaments-evidence that the substantive effect of containment was slow to take effect (Chadha, 2015, 109). Furthermore, as A. Hashim notes, the denial of supplies of heavy weaponry (the LTTE’s main foreign dependence) did not preclude the adoption of a guerilla strategy utilizing more locally available small arms (Hashim, 2013, 128). The impact on Prabhakaran’s decision-making, however, was disproportionate. The increasing desperation of Prabhakaran is perhaps best demonstrated by his erratic response to the closing financial net- a series of terrorist attacks on both Sri Lankan political and military figures and (catastrophically) a 2006 attack on several foreign diplomats- actions which only served to entrench the popular image of the LTTE as a terrorist outfit (Hashim, 2013, 24). In a further act of erratic decision-making, the LTTE chose in 2006 to close the sluice gates of a vital dam in the Mavil Aru district to the agricultural communities below-providing the administration of president Rajapakse , in the president’s own words, with a political “green light” to resume hostilities (Mehta, 2010, 6). Once fighting began, Prabhakaran showed an almost obsessive desire to concentrate his forces at key points such as Killonchchi- even going so far as to deploy the remnants of the Sea Tigers as ground forces- intimating a lack of concern for the long-term maintenance of supply lines and the assumption that the war would be ended quickly and decisively. This stands in sharp contrast to Eelam war III where the LTTE proved willing to surrender key bases such as Jaffna, whilst resorting to asymmetrical strategies of mass suicide bombing against Sri Lankan forces, coupled with decisive assaults on the exposed rear of the Sri Lankan army at Mullaitivu (Pape, Feldman, 2010, 300). Nor was this entirely a strategy forced upon Prabhakaran and the LTTE. Indeed, prior to the onset of hostilities, a report prepared for the administration of president Rajapakse by the Sri Lankan joint chiefs of staff noted significant concentrations of forces by the LTTE-intimating that had Prabhakaran not provoked a response he intended to carry out a conventional assault beyond the confines of his territory- a gambit even more risky than the one he eventually took (Mehta, 2010, 100). It is unclear, then, why Prabhakaran chose to abandon a strategy that, as several scholars have noted, delivered decisive success (DeSilva-Ranasinghe, 2010, 6) (Jackson, 2007 ,67-70).
An insight into the roots of this shift is provided by the LTTE’s propaganda, however. Specifically, it has been noted that through the decade, the LTTE’s appeals to overseas financiers took the form of promising a decisive “final war”. Notably, the occurrence of this rhetoric covaries with the LTTE’s increased isolation. Indeed, in his 2006 Heroes Day speech, Prabhakaran explicitly made an appeal to the Tamil diaspora, noting that contributions would become increasingly difficult to make, but promising that a last push to equip the LTTE would make possible a decisive end to the conflict. Notably, Prabhakaran made specific allusions to the GOSL’s “racist Sinhala elements” misuse of the ceasefire as an opportunity to weaken the LTTE as evidence both of the need for renewed hostilities and a decisive victory (Murari, 2012, 100). Effectively, then, Prabhakarans rhetoric directly conforms to that of an actor in a frame of losses, facing what he saw as the high probability of debilitation due to containment or the low possibility of success via a decisive defeat of the SLAF (a risk acceptant option which, if it failed, would signal the immediate end of his organization). True to the predictions of prospect theory, Prabhakaran chose the risk acceptant strategy.
However, an explication of the GOSL’s success in putting Prabhakaran in a frame of losses through the lens of the ceasefire and financial containment overseas would be incomplete. A second part of the GOSL’s strategy aimed at the Sun Tzunian policy of attacking Prabhakaran’s intra elite alliances-thereby securing political support from the Tamil community and placing Prabhakaran further in a frame of losses. Specifically, as noted by Colonel U. Pareira, the LTTE was naturally divided along geographical lines, insofar as the insurgents in the east were drawn from backgrounds that were economically vastly different from the more urbane recruits of Prabhakarans urbane northern support base (the eastern Tamils had largely migrated to the island as plantation workers) (Pareira, 2012, 55). While these divisions had been nascent within the organization, the consistent conflict with the Sri Lankan army that characterized the 1990’s largely induced in group cohesion out of necessity (DeVotta, 2010, 165). The tactical ceasefire, however, removed this source of in group cohesion. Furthermore, the Sri Lankan Army contributed to the nascent split by the articulation of its policy of containment at the operational level in the early years of the decade. Specifically, the Sri Lankan government fell back upon a string of fortified bases that it had constructed between the north and east in the 1990’s. While the kinetic strategy of the 1990’s had rendered these bases liabilities, the more static policy of the first half of the next decade effectively ensured that the Sri Lankan army straddled all the key routes between the north and the east- most crucial among which was the A11 highway. Furthermore, the movements of supplies (particularly military supplies) by sea between the east and north was problematized by the terms of the ceasefire and Sri Lankas patrolling of sea lanes (Marks, 2007, 51-54). This had the dual effect of ensuring that the LTTE could not effectively maintain either centralized control or a steady redistribution of resources to its eastern faction. Consequently, the increasingly popular leader of the eastern faction Colonel Karuna saw his grievances against the north multiply even as his prospects for a successful break from the LTTE seemed to be enhanced by the cordon sanitaire between him and Prabhakaran’s wrath (Hashim, 2013, 300). This was quickly noted by the then Sri Lankan government of C. Kumaratunga, which seems to have played a key role in offering Karuna the assurances that he needed to defect, which he did in 2004. While President Kumaratunga herself strenuously denies this, the fact that Douglas Devananda the leader of the EPDP (which in turn was a member of Kumaratunga’s coalition) played a pivotal role in integrating Karuna into civilian politics following his defection seems to imply that Kumaratunga’s protestations were calibrated towards the sentiments of her domestic audience (Mukarji, 2005, 29). To be sure, the seeds of the split existed independent of the Sri Lankan governments actions but, as Z. Mamphily points out, the centrality of the lack of resource allocation to Karuna’s defection (itself a function of GOSL strategy) is the central qualm cited by leaders of the defecting faction (Mamphily, 2011, 222). At the very least, the GOSL provided the occasion, if not the cause, for the split.
The policy of attacking intra elite alliances, much like that of containment writ large, is notable not for its material effect on Prabhakaran but its impact on the way in which he framed the conflict. As Rabasa et al, commenting at the time noted, Karuna’s force was not large enough to represent a debilitating loss of manpower for the LTTE. This was particularly true as the LTTE’s largest units the Black Tigers and the Leopards division were both based in the north and extremely loyal to Prabhakaran (Rabasa et al, 2006, 26). Indeed, even in the east Prabhakaran retained sufficient numbers of loyal troops to decimate Karuna’s forces in a series of pitched battles (Dissanayaka, 2005, 271). Rather, the impact of the defection seems to have been overwhelmingly psychological. As Swamy (2010, 253) notes, dissent in the east was met with frenetic attempts at centralization, including the assassination of rival leaders within the LTTE – a policy which only served to weaken Prabhakarans hold on the east. Simultaneously, as noted by C. Clarke, the defection of Karuna seems to have coincided directly with efforts to concentrate forces in the east in strongholds such as Thoppigala and Batticoloa to meet the needs of a conventional warfare strategy (Clarke, 2015, 100). This strategy is a far cry from the what might be dubbed a Phase I Maoist strategy that had been followed in the east, which centered around small decentralized guerillas that exerted influence on urban areas only through infiltration and informal “taxation” by plainclothes agents (Gunaratna, 1997, 98). It appears, prima facie, that the Sri Lankan governments success in driving a wedge between the two regional centers of the LTTE placed the LTTE’s leadership in a frame of losses-whereby it felt its ability to sustain its movement in the long term was eroding. Crucially, this coincides with the adoption of significantly more risk acceptant strategies by the LTTE in the east- a fat that meant that when they confronted the Sri Lankan army it was as a conventional force, with predictable results. Most importantly, however, the impact on the LTTE’s capacity to control the east was more perceptual than real and mattered insofar as it served to shape the LTTE’s strategy in the forthcoming war.
Finally, we must turn to the operational strategies adopted by the Sri Lankan army. Even if prospect theory explains why in the buildup to Eelam War IV Prabhakaran adopted a suboptimal strategy, it is unclear why he did not recalibrate his strategy after initial military setbacks. At this point, I argue that a useful corrective is Colonel John Boyd’s celebrated theory of the Observe, Orient, Decide, Act loop. Specifically, Boyd argued that any actors reaction to a changing scenario during military actions can be disaggregated into these fairly self-explanatory components. Consequently, Boyd contended that an operational policy that centered around pace and the use of multiple fronts rendered both orientation and reaction problematic (Osinga, 2005, 2). This effectively represents the synthesis of principles articulated by, amongst others, Liddell Hart, who stressed the importance both of calculated dispersion and obscuring one’s possible axis of advance (Liddell-Hart, 1991, 94). Consequently, Boyd made a case for maneuver centric war along multiple fronts that effectively paralyzed an opponent’s reaction capacity. Typically, as R. Jervis notes, when an actor’s capacity for perception and reaction is hobbled by uncertainty and the pace at which a situation changes, it reflexively adheres to its original plans (Jervis,1976,195). Consequently, I argue that by examining Sri Lanka’s operational strategy during Eelam War IV, we can understand how a conventional force could ensure that the gains of grand strategy (in terms of compelling its insurgent foe to fight with a direct defense) were retained. Below, I will discuss the military reforms that allowed the Sri Lankan army to operate within the LLTE’s decision-making loop, along with the operational strategy that put these reforms to use.
The GOSL utilized the ceasefire period to make several reforms to its military force structure and doctrine. Firstly, the creation of the paramilitary CDF (Ceylon Defence Force) to secure the army’s rear ensured that the Sri Lankan government would not have to attenuate its military forces due to garrisoning duties or leave itself vulnerable to attacks on its rear as had occurred during Eelam War III (Rajapakse, 2010, 1). Simultaneously, the army was increased in size, with recruitment in 2006 amounting to 36,000 fresh troops (DaSilva, 2009, 1). Most vital, however, were changes to the Sri Lankan army’s force structure. Noting the difficulties faced by large bodies of troops in clearing the heavily forested terrain of the island, infantry reformers such as then Major General Sarath Fonseka pushed for the induction of 12-man SIOT’s (special infantry operating teams) to operate in conjunction with regular forces. As Fonseka noted, the ability of the Sri Lankan army to insert SIOT’s on the flanks and rear of the LTTE’s positions induced a degree of uncertainty into the LTTE’s commanders- inducing them to adopt the tactical expedient of positioning all of their forces behind fortified bunds (trenches and mounds)- leaving themselves vulnerable to the firepower of regular forces (Glenn, 2015, 28). By 2006, the Sri Lankan army had 6,000 troops serving in SIOTs (almost treble the number it had a decade earlier). With the appointment of Fonseka as Commandant of the Sri Lankan Army by M. Rajapakse’s government midway through the decade, this transformation would intensify. The impact of these efforts was to create a force that was simultaneously capable of concentrating large levels of mass and achieving significant tactical and operational mobility. Simultaneously, as has been noted earlier, the Sri Lankan navy underwent significant doctrinal changes under Vice Admiral Karannagods, coupled with the introduction of indigenous small boats in line with the eponymous small boats concept which allowed the navy to counter the Sea Tigers swarm tactics. While the significance of this change at the grand strategic level has already been discussed, it also bore relevance to the Sri Lankan operational strategy. Specifically, control of the sea made it possible for the Sri Lankan navy to conduct amphibious landings and provide sea-based fire to troops operating on the coast. To this end, several indigenous fiberglass landing craft were added to the navy’s arsenal in the war. The impact of this naval overhaul was, then , to enhance the mobility of Sri Lankan forces which could be landed to the rear or flanks of opposing LTTE formations with a rapidity that has been noted by sea power theorists since Mahan (Mehta, 2010, 10) (Mahan, 1918 ,100).
Following Fullers injunction, however, these changes represented the foundations of an operational strategy but were not the strategy itself. The operational strategy enabled by these changes effectively centered around three tenets. Firstly, the objective of the campaigns of Eelam War IV was the LTTE’s armed forces rather than seizing and holding territory, as traditional COIN theory would advise. Secondly, operations in different theaters would be carried out either simultaneously or in quick succession. Finally, within each individual theater, the Sri Lankan army would advance upon multiple fronts, attempting where possible to choose axes of advance to the flanks and rear of the LTTE’s forces, thereby unbalancing the LTTE’s reaction process by leaving its commanders, as William Tecumseh Sherman had it, “on the horns of a dilemma”. Consequently, the eastern campaign began with the Sri Lankan army, which had broken the Tigers siege of mavil aru, turning the flank of the LTTE’s position in the east with an amphibious landing in Muttur (which was controlled by a s mall detachment of SLA forces) supported by naval gunfire, allowing it to advance to and capture the major base at Sampur, inflicting heavy casualties on the LTTE along the way. Simultaneously, SLA forces advanced overland to LTTE bases at Thoppigala and Batticoloa. At the tactical level, the use of SIOT’s guided by intelligence furnished by Colonel Karuna allowed the SLA to target the lines of communication and supply to each base. The conjunction of multiple simultaneous advances and the use of SIOT’s at the tactical level left the LTTE unclear as to the main objective of the SLA and induced its commanders to seek repose in the strategy that they had pursued thus far-concentrating forces near strongpoints in the hope of inflicting a decisive defeat on one of the Sri Lankan forces as a prelude to negotiations-with predictably fatal results.
The eastern campaign was followed up in short order by an advance into the northern theatre of the war-indeed the northern offensive coincided with the fall of Thoppigala in July 2007. Crucially, while the Sri Lankan government had made efforts to stress the need to carry out the operations in rapid succession, by the time the offensive on Prabhakaran’s holdings in the north had begun, the LTTE had suffered a string of defeats that might induce him to recalibrate his strategy. However, consistent with the predictions of prospect theory, Prabhakaran’s aides state that their leader adhered to the hope that a decisive victory would render the GOSL’s campaign politically unviable (Singh, 2010, 52). Furthermore, consistent with the broader operational principles it had adopted, the GOSL’s northern assault was geared towards operating within the LTTE’s decision-making loop and paralyzing its capacity for recalibration. Specifically, the army advanced across multiple fronts, threatening multiple objectives simultaneously. Consequently, it was not clear until late in the campaign that the armys main thrusts were against Mannar to the west and Mullaitivu in the east (Mehta, 2010, 12). In addition to the main thrust of the offensive along the A32 highway, which served to link Jaffna to the mainland, the Sri Lankan army launched attacks from Jaffna in the north along the Kandy Jaffna highway, while columns moved against Omanthai from the south- advances that pinned the LTTE to Killonchchi by rendering an attack on this stronghold plausible. Simultaneously, the 59thinfantry division moved against Mullaitivu from Wali Oya in the south (Mehta, 2010, 8). At the same time, SIOT’s backed by airpower effected deep penetrations along the road links that served al LTTE lines of communication-effecting the paralysis cited as central by RMA theorists who, in no small part, are inspired by Boyd. As one analyst put it, the inability of the LTTE to perceive the main axis of advance led the LTTE to concentrate its forces in Killonchchi and Mullaitivu, even as the Sri Lankan government dispersed its forces- leading to a strange inversion of roles where the insurgent was more static than the conventional force (Mehta,2010,8-12). This behavior is in keeping with organizational theorists who stipulate that groups tend to adhere scrupulously to “plan A” to make sense of a rapidly changing situation. In Prabhakarans case, this meant adhering to a fatal policy of concentrating large formations of forces at key points. This concentration of forces along with leadership assets left the LTTE vulnerable to rapid disintegration in the event that these forces were destroyed, as indeed they were when Mullaitivu and Killonchchi, both left surrounded by the initial SLA thrusts, fell in the wake of the dual pincer movements on these strongholds that constituted the final act of the campaign (Bandarage,2009,50-52). Indeed, Parabhakaran is later said to have confided in his intelligence chief Pottu Amman that over 75 percent of the LTTE’s strength had been concentrated in the hope of defending these assets, and that the loss of this force had effectively doomed the movement. (Mehta, 2010, 12).
Importantly, this approach stands in sharp contrast to the third Eelam war, which saw the LTTE, under Prabhakarans leadership, abandon its prized Jaffna base to prosecute attacks on the Sri Lankan army’s rear at elephant pass and Mullaitivu (Manage, 2010, 18). If the LTTE adopted a flawed conventional strategy, then, this can be attributed to the Sri Lankan governments grand strategy of containment, coupled with its operational strategy of operating within Prabhakaran’s OODA loop by operating across multiple fronts, with multiple prospective objectives, at a very high tempo.
Conclusion and Implications
The successful application of prospect theory to insurgent behavior demonstrates that the selection of a strategy by an insurgent is not exogenously given, as much counterinsurgency presumes as a starting point, but is open to being shaped by the counterinsurgent state. This allows us to add a progressive hypothesis to the theory of asymmetrical warfare developed by scholars such as Toft (2005) who contends that a same strategy interaction (in which both actors adopt a direct strategy aimed at each other’s forces) is likely to result in a short, decisive wars while opposite strategy interactions protract a conflict. Building on the insights offered by Kahneman and Tversky, I hope to have proven that by the apposite application of diplomatic and political pressure, the counterinsurgent can induce the insurgent to risk acceptant behavior, leading it to adopt a direct strategy and effect a short decisive interaction.
A major implication of this would be the implied reversion of the commonplace doctrine of “clear, hold, build” that characterizes territory and population-centric theories of counterinsurgency. Indeed, by eschewing this doctrine (which it had previously practiced) the GOSL caused the insurgency to coagulate into something approximating to a conventional force even as it fractured internally. This was achieved not by clearing territory but by temporarily eschewing control over it and effecting an increasingly effective policy of containment and isolation both locally (by controlling sea lanes and key highways) and internationally (by linking its war to the broader war on terror). Additionally, the counterinsurgency strategy of Eelam war IV explicitly eschewed population centric theories of counterinsurgency, focusing instead on an eminently Clausewitzian enemy-centric approach that threated the insurgents armed forces as his center of gravity.
At the operational level, the key implication is that the theories of Boyd, long incorporated into conventional warfare, are applicable and even central to the destruction of an opponent within an enemy centric context. By attacking on multiple fronts, the GOSL effectively precluded any within conflict recalibration of strategy by Prabhakaran- ensuring that the strategic gains of placing the LTTE in a frame of losses did not wither away. The destruction of the LTTE’s concentrated forces, then, followed a pattern familiar to any conventional engagement between two mismatched foes- a remarkable achievement for a counterinsurgency.
Collectively, then, the implication of this study would be the potential for “clear, hold, build” to be replaced with a strategy of retrenching, containing and then effecting military destruction. In this paradigm the enemy, rather than space or population, becomes salient in a manner familiar to traditional practitioners of war.
Finally, I note that there remains room for further research on the topic in order to improve the generalizability of my findings. That said, the variation of the LTTE’s strategy between the first three Eelam wars and the fourth even as Prabhakaran remained at the organizations helm should allow us to eliminate the alternative hypotheses built on group and leader idiosyncrasies. Furthermore, I hope to have eliminated alternative hypotheses built around insurgent legitimacy and state strength earlier in the paper.
Conclusively, then, the application of prospect theory and Boyd’s OODA loop offer rich insights into the existence of alternative ways of prosecuting a COIN strategy.
No comments:
Post a Comment