8
Implications

To wrap up the workshop, planning committee member David Laitin of Stanford University described what he saw as the four main themes, or focuses, that had emerged during the two days of presentations and then offered several “lessons learned.”

WORKSHOP THEMES

To introduce the four themes, Laitin offered a thought experiment. Suppose that the U.S. military is planning a major occupation of Somalia in order to fight the insurgency led by the Islamist group al-Shabaab, which is increasingly linked to Al Qaeda. The goal of the occupation is to provide support to the transitional federal government of Somalia. At the moment the transitional government is in control of only a small section of the entire country—part of Mogadishu, the country’s capital, and nothing else—with help from troops from the African Union, mainly from Uganda and Burundi. It is a situation, like the one in Iraq and Afghanistan, that will require more than military firepower to succeed.

With this in mind, General James Mattis, U.S. Marine Corps, commander of the U.S. Joint Forces Command,1 poses the question, What cultural knowledge do we need in order to maximize our chances of success?

1

General Mattis was commander of the U.S. Joint Forces Command 2007-2010, after which he became commander of U.S. Central Command.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 83
8 Implications T o wrap up the workshop, planning committee member David Laitin of Stanford University described what he saw as the four main themes, or focuses, that had emerged during the two days of presentations and then offered several “lessons learned.” WORKSHOP THEMES To introduce the four themes, Laitin offered a thought experiment. Suppose that the U.S. military is planning a major occupation of Soma- lia in order to fight the insurgency led by the Islamist group al-Shabaab, which is increasingly linked to Al Qaeda. The goal of the occupation is to provide support to the transitional federal government of Somalia. At the moment the transitional government is in control of only a small section of the entire country—part of Mogadishu, the country’s capi- tal, and nothing else—with help from troops from the African Union, mainly from Uganda and Burundi. It is a situation, like the one in Iraq and Afghanistan, that will require more than military firepower to succeed. With this in mind, General James Mattis, U.S. Marine Corps, com- mander of the U.S. Joint Forces Command,1 poses the question, What cul- tural knowledge do we need in order to maximize our chances of success? 1 General Mattis was commander of the U.S. Joint Forces Command 2007-2010, after which he became commander of U.S. Central Command. 83

OCR for page 83
84 SOCIOCULTURAL DATA TO ACCOMPLISH DOD MISSIONS And that is the approach Laitin said he used to identify the key messages from the workshop: “What have I learned in the past two days that will help me give General Mattis a state-of-the-art answer?” The value of Stories Continuing with his hypothetical scenario, Laitin said the first part of the answer is that “the new commander of the International Security Assistance Force–Somalia should hear some stories.” Stories are a good starting point for understanding a situation and for generating ideas about how best to deal with it. They also serve as a basic source for the analytical hierarchy process and the Delphi method that domain experts can use in solving various sorts of problems, as Robert Sargent described in his workshop paper, “A Perspective on Modeling, Data, and Knowledge.” What sorts of stories? Some of them should be from people experi- enced in past campaigns, Laitin said. Robert Oakley, the special envoy for Somalia under President George H.W. Bush, could describe his suc- cessful management of Operation Restore Hope from 1992 to 1994. “He saved ten thousand lives getting food out to Baidoa and other cities,” Laitin said. Another story worth hearing would be that of Mahmoud Sahnoun, a United Nations special envoy who worked with the various warring factions in an effort that might have stopped Somalia’s civil war if he had received support from some of the world’s powerful countries. Then there was Colonel Kenneth Allard, U.S. Army (retired) who wrote Somalia Operations: Lessons Learned (2002), a book that described how the UNOSOM II mission to Somalia collapsed, which led to the disastrous Blackhawk Down battle. There are also a variety of local stories to be told, Laitin said. One would describe how the Hawiye clan broke up, which was the root of the collapse of Somalia and the resulting 20-year war. Another would be the tale of how the Isaaqs in Hargeisa were able to negotiate a settle- ment with the Warsangeli and Dhulbanhante Daarood, an act that has allowed Somaliland, an autonomous region of Somalia, to have main- tained a long-standing peace amidst the chaos that reigns in the rest of Somalia. And what were U.S. policy makers thinking when they encour- aged Ethiopian troops to invade Somalia in an attempt to overthrow the Islamic Courts Union regime? A story could illuminate the reasoning behind that decision. “Stories are a basic source for analytic hierarchy process of domain experts, and I think these stories are essential and comprehensible,” Laitin said, “and that is what makes them essential for any commander.” They

OCR for page 83
85 IMPLICATIONS are also, however, necessarily incomplete, no matter how compelling they may be and any military officer should keep that in mind. Data-Driven Target Identification The second theme to emerge from the workshop that might be of use in a hypothetical mission to Somalia, Laitin said, was the idea of “data- driven target identification,” or accumulating information to identify enemies (e.g., terrorists, violent extremists, criminals). The presentation by David Kennedy described one approach to such identification. His group finds core offenders by interviewing both the local police and gang members and then, once those most likely to be violent have been identi- fied, uses proactive community policing to deter them from killing others. Hsinchun Chen offered a very different approach, employing computers to analyze close to five billion pages, files, and messages from the Dark Web in order to find terrorists and potential terrorists and determine their relationships with one another. The two approaches, while different, have a number of commonali- ties. In both situations, the core offenders are relatively few but are the cause of most of the violence. At the same time, however, violence is generally a product of groups of individuals rather than the individuals themselves. The groups are fluid and dynamic, so Laitin indicated that it is necessary to follow them carefully as they develop and change. Finally, Kennedy indicated that the key to controlling the violence is to gather local information on both the targets of the violence and on the people who can sanction those committing the violence. That is, the best way to control the violence is to change the social dynamic at a local level rather than exerting outside pressure via the police and the court system. Are there any insights from this approach that might aid the theoreti- cal commander of the International Security Assistance Force–Somalia? Laitin suggested that there are. The standard approach to understanding Somali group structure, going back to the work of social anthropologist E.E. Evans-Pritchard in the mid-20th century, is to focus on segmented lineages, and this is the approach taken by the National Counterter- rorism Center operation in Djibouti, which borders Somalia, Ethiopia, and Eritrea. “However,” Laitin, said, “these segmented lineages have in many ways collapsed, and other groups have formed.” That is not clear from reading the standard anthropology literature, he continued, but “we would know it from either Chen’s or Kennedy’s work, which would force local people to ask, Where are the groups, how are they formed, and how do they interact?”

OCR for page 83
86 SOCIOCULTURAL DATA TO ACCOMPLISH DOD MISSIONS He noted that, although the Chen and Kennedy approach of identi- fying the core offenders and studying group relationships would offer insights into where the violence is coming from, they do not provide much guidance on how to keep the violence in check. “Policing in Afghanistan or Somalia is not like policing in Cincinnati,” Laitin noted. Cincinnati is part of a larger, well-ordered nation in which gangs are not a risk to hide out in the bush and become insurgencies, and so the violence remains local. As Robert Albro noted in an earlier discussion, the U.S. military is handed a much bigger job than simply finding and controlling problem- atic individuals. Thus while Kennedy’s community policing approach may be effective in lowering the murder rate in an American city, it may not offer much of a blueprint for creating a state that will be able to con- tain its own violent offenders once the Americans leave. Cultural Models The third theme that might be of use in a hypothetical mission to Somalia is the use of cultural models. It is important to keep in mind what such models can do, Laitin said. “Can cultural theory predict indi- vidual behavior? [Robert] Rubinstein is right when he says no, but that’s an unfair question, because predicting individual behavior is an absurd requirement of any theory.” What cultural models can do, however, is to point to tendencies or probabilities, and that should be the standard by which they are judged. The weakness of all of the cultural models described at the workshop, Laitin said, was that they offered no “engineering,” that is, no specific recommendations for how to train soldiers to win the hearts and minds of the civilian population. “The Michigan team on psychological orientation of the WEIRDos [Western, educated, industrialized, rich, and democratic cultures] versus the rest is really world-class research, but there’s no direct implication other than making people culturally aware that Asians are dif- ferent from us WEIRDos (Henrich et al., 2010). Similarly applying Jeanne Brett’s distinction between honor and face cultures—those distinctions are clear and meaningful, but the implications for what you do in a train- ing program and how they’ll make soldiers better soldiers is completely unspecified.” So it would make no sense, Laitin said, to offer these cul- tural models in military training as if they had operational significance. Indeed, he noted, when the speakers who presented the cultural models were asked about how they might be applied in military training to help troops headed to Afghanistan or Somalia, the presenters tended to fall back on traditional ethnography: “Well they’ve got to learn the language, and they’ve got to learn the basic social structure and all this— disconnected, as it were, from the models they just presented.” Sugges-

OCR for page 83
87 IMPLICATIONS tions that soldiers should be familiar with local customs or learn the local language are nothing new, Laitin commented. Still, he continued, understanding the local culture can be useful in a number of ways, such as offering insights into political maneuvering. Latin offered as an example an episode from recent Somali history. “The dictator Siad Barrée was in power from 1969 to 1990, and after he lost the war against Ethiopia in 1978, most of the tribes other than his own clan turned against him. He was fighting three insurgencies, and he had to worry about internal coups of people trying to overthrow him from within.” So he chose to appoint General Mohamed Ali Samantar as his sec- ond in command. Samantar was a clever choice, Laitin explained, because he was an outcast in the Somali clan structure, a man who would have faced country-wide rebellion if he ever took over Barrée’s job. “Samantar understood this as well, which meant he had no incentive to overthrow the dictator, and so the dictator got free protection from someone who liked being No. 2, but could never be No. 1.” Understanding a local culture can be thought of as a way of tapping into what is common knowledge in that area—such as the knowledge that Samantar would never be a threat to Barrée’s power. In particular, Laitin said, cultural knowledge helps one understand coalition dynamics, which is a very powerful tool. It is the sort of tool that Kerry Patton was talk- ing about in his presentation calling for a new discipline of sociocultural intelligence, or SOCINT. Such a discipline might help interpret practical politics in foreign countries by bringing to bear an understanding of social relationships and how the local population interprets coalitions and other political phenomena. This might be particularly useful in Iraq, Laitin said, as it is “a country that hasn’t had a government in a few months, because they can’t form a coalition.” Laitin also addressed two questions from the audience having to do with cultural theory and how to train military ethnographers. One ques- tioner pointed out that there is a great deal of sociocultural data available, but they are not organized in such a way to be easily findable and usable. Would it make more sense to work to organize these data or to collect new data each time information on a culture is needed? Laitin suggested that it’s not practical to prepare for every culture in the world in a place that might become important to the military. Instead, it makes more sense to have a corps of specialists trained in ethnography who can quickly collect and interpret data on places once they have become important. The second audience question asked was how much time it would take to train soldiers to the point that they are adept in another culture. It would be a long-term process, the questioner observed, maybe years long. Laitin responded that one way to deal with this reality would be to develop an ethnographic specialty in Army training so that an elite group

OCR for page 83
88 SOCIOCULTURAL DATA TO ACCOMPLISH DOD MISSIONS of soldiers knows how to collect useful cultural information and to train others in data collection. Finally, Laitin pointed to the flattening of the command structure described by MG Flynn as an important way of spreading ethnographic understanding through the military force in an area. Lower level officers and the noncommissioned officers are already contributing to a Wiki knowledge base with information on what works and what doesn’t work in dealing with the local population, details about the local power struc- tures and social structures, and other sociocultural information that may be useful. This is very similar to the approach that Kennedy described of using police as local experts on what is happening on the streets of Cincin- nati, and it could well prove to be equally effective. Systems Models The final theme that Laitin discussed was the idea of systems models. Can systems models provide the unifying answer that the military would like? “Captain Schmorrow wants products that are not specific to a single decision, but something more general,” Laitin said. At the same time, how- ever, Captain Schmorrow said the use of models could not be practically limited to general theorizing and situational understanding and should be able to make decisions in specific cases. Decision makers will use what- ever information is available to make decisions, especially when working within extremely short timelines. Captain Schmorrow said, referring to a hypothetical situation, “most decision makers have about 30 seconds or three minutes to say, ‘Did the model say to dig a well or to give them soccer balls?’” He explained that the unifying answer the military is looking for is to deliver messages that accurately reflect the U.S. position and resonate with the way that other groups or societies understand the world. It is certainly the case that some systems models, such as agent-based models or systems dynamics, are unifying, Laitin said, so the real question is whether these models can deliver answers to the sorts of specific ques- tions that the military faces in the field. “It’s hard to answer that,” he said, “since none of them was described or defended. We got only critiques of them and suggestions of what they might do, but we actually didn’t have any examples presented of them.” More specifically, he said, a number of questions surrounding these models have not yet been answered. First, there is no evidence, for exam- ple, that the models can make predictions at the success rates that some claim are possible. Some of the claims made for them are “beyond belief,” he said, and are unsupported by any evidence. Second, there is no evidence that the models can demonstrate emergence—that is, the appearance of phenomena that have not already

OCR for page 83
89 IMPLICATIONS been programmed into them. The example described by Laura McNamara (whereby agent-based modeling is used to look at emergent behaviors in water sharing and resource management in Bali—Lansing, 1994) might be an exception, Laitin said, but it is important to show not only that a model generates lots of information but also that one is actually getting more out of the model than was put into it. “If the information that you’re feeding your agents is more or less what you’re getting out of it, we would say there’s no emergence and we’re not doing much.” Third, there is no evidence that models can aggregate local knowl- edge, at least not in a way that makes it more understandable or compre- hensible to decision makers or to operations people in the field. Not only that, but as planning committee member Catherine Tinsley pointed out, the models may actually have slippage, that is, lose information rather than generate new information. Fourth, there is no evidence that systems models are any better at demonstrating causes than empirical models. One possible exception is work done over the past 15 years in econometrics. “There’s a whole range of econometric tools that get at causes in new ways,” Laitin said. However, although a number of economists were invited to speak at the workshop, none of them was able to attend. Finally, and most importantly, there is as yet no evidence that these models can address the sorts of questions that a commander in the field would have. Indeed, it is not even clear that they can provide the “pos- sibility space” for the sorts of problems a commander would like to address. If models cannot answer pointed questions, Laitin said, or if they are not designed to answer these sorts of questions, then it is reasonable to ask if it makes sense to continue developing them. LESSONS LEARNED To finish his presentation, Laitin offered a series of lessons learned. These lessons represented his personal opinions rather than a summary of what he heard from the workshop presenters. The first lesson, he said, is that “counterterrorism, as I’ve been con- vinced over the past couple of days, requires careful police work.” Fur- thermore, Laitin said, comments from Andrew Imada and David Kennedy lead to the conclusion that since the culture and the training of the U.S. military are designed for missions in which a known enemy is located and neutralized, “the military may not be the correct organization culturally to be able to do the kinds of things necessary” in missions like the one in Afghanistan. The second lesson, Laitin said, is that basic science seems to get short shrift in this area. In particular, there are few serious field evaluations of

OCR for page 83
90 SOCIOCULTURAL DATA TO ACCOMPLISH DOD MISSIONS new proposals, programs, or models. Not even the Human Terrain Teams or the various cultural training programs, which are already in the field, are being subjected to any kind of field test. “We do no scientific evalua- tion that I can see of our own initiatives.” A related issue is how to involve cultural anthropologists in this area in a way that does not violate their professional ethics or norms. In World War II, Laitin noted, the British historian George Taylor assembled a group of top anthropologists to answer specific questions about fighting the Japanese, such as why their surrender rates were so low. “One of the things that Taylor understood was that you can’t bring anthropologists to help choose targets,” Laitin said. He believes the United States has not fully applied Taylor’s lessons in harnessing the best scientific evidence for the fulfillment of its field objectives. A third lesson is the importance of the group as a fundamental unit of cultural understanding. This was apparent in a number of presenta- tions, Laitin noted, including those of Chen and Kennedy as well as Brant Burleson’s talk on social support. “Understanding radical group dynam- ics, what makes them efficient, what makes them resistant to negotiations, what makes them murderous,” to Laitin, seems like “areas of research that may benefit the U.S. military.” Finally, Laitin repeated the words of Robert Rubinstein in his presen- tation on culture in cooperative relationships that trying to create some sort of general predictive model of social and cultural influence on behav- ior is likely to be a “fool’s errand.” Laitin echoed the general sentiment of several of the workshop presenters: attempts to create broad, integrated approaches to social science issues or to base practical applications on such integrated theoretical foundations are not likely to be successful. The more valuable approach, he said, would be to work to extract local information on targets, the authorities and informal leaders, and cultural practices— what MG Flynn referred to as “the environment”—in order to address specific field questions. Throughout the workshop, participants seemed to have at least one position in common: although models are not the single answer to the sociocultural challenges faced by the U.S. military, they may have utility in strategic-level sense-making and providing a “probability space” to assist commanders in making better informed decisions.