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 17
Sociocultural Data to Accomplish Department of Defense Missions: Toward a Unified Social Framework - Workshop Summary 3 The Conflict Environment As has always been the case, the main focus of the U.S. military today is to carry out successful operations in environments of conflict, but the types of conflict environment in which the military operates now are dramatically different from what has long been the norm. In place of battlefield combat with distinct front lines, often taking place in the countryside, today’s armed forces are more likely to carry out their operations in urban areas or villages with few clear boundaries. Similarly, the lines between combatants and civilians have become blurred, as the combatants are often dispersed throughout the civilian populations, and a civilian today may be a combatant tomorrow. Because of these changes, sociocultural knowledge has become an increasingly important factor in the success of military missions in conflict environments. The workshop’s first panel addressed ways in which a sociocultural approach can offer new perspectives on conflict and violence that may improve the chances of success in such missions as combat, counterinsurgency, and counterterrorism operations; collecting and securing light arms and military weapons; protecting local noncombatant populations; and militia and criminal gang suppression. To this end, speaker Hsinchun Chen described methods he has developed to identify “dark networks” of criminals and terrorist groups; David Kennedy detailed ways to significantly reduce homicide rates in large cities by identifying and focusing on the groups most likely to be involved; and Kerry Patton described how both the military and the intelligence
OCR for page 18
Sociocultural Data to Accomplish Department of Defense Missions: Toward a Unified Social Framework - Workshop Summary community could benefit from the establishment of a new discipline of sociocultural intelligence. MAPPING THE DARK WEB When dealing with terrorists, one of the most difficult problems to overcome is that they are mostly hidden, so that it is difficult to identify them and follow what they are doing. In his presentation, Hsinchun Chen of the University of Arizona described how he has used information from the Internet—forums, chat rooms, video postings, and so on—to identify potential terrorists and to map out their web of connections. Beginning in the late 1990s, Chen developed a system called COPLINK that provides a way to link information from a large number of criminal justice databases, such as collections of detailed criminal reports from local police departments (for additional background information, see Hauck et al., 2002). He also developed a variety of software tools for analyzing the collected data, such as one called COPLINK Detect, which searches for the presence or absence of links among people, vehicles, places, and offenses in the police reports, allowing investigators to notice connections that might otherwise be overlooked. After his work on COPLINK, Chen said, he became interested in the area of international security when he read the book Understanding Terror Networks by Mark Sageman (2004). “He was making the claim—and also providing the data—that most international jihadists are using the Internet as a recruiting, assignment, training, and communication tool,” Chen said. “He says this is leading into something called leaderless jihad.” This description of the terrorist world convinced Chen that he could develop tools to help in this area as well. His academic background is in computer science, Chen explained, and specifically in the discipline of intelligence and security informatics, which he described as using information technology systems, databases, software models, and other computer-based tools to deal with security-related issues (see Chen, 2006). Although COPLINK is not directly related to the work he has been doing on terrorist networks, the two share a common approach: that of combing through vast amounts of information looking for understanding and insights. When he first started work on his new project, which he called Dark Web to refer to people who were present on the Internet but trying to remain hidden or “in the dark,” his goal was to “collect all the terrorist-generated content in the world.” He looked at Aryan Nation, the Ku Klux Klan, and other domestic groups in the United States. He looked at Arabic-speaking groups and Spanish-speaking groups. Eventually, however, he decided to focus mainly on jihadist groups, Islamic-inspired groups, and
OCR for page 19
Sociocultural Data to Accomplish Department of Defense Missions: Toward a Unified Social Framework - Workshop Summary individuals expressing interest in holy war. He has developed a multistep process to collect and analyze data from the sites that such groups host or visit (see Figure 3-1). The first step, Chen said, is to identify as many places as possible on the Internet that have a potential connection with Islamic extremism and terrorism. “We started out collecting websites, and we got into forums and blogs and YouTube.” The people who are putting this information on the web are interested in reaching new listeners, so it is generally straight-forward to sign up and get access to the sites. Then his group starts to collect information from the sites using web crawlers that go from page to page, gathering information from each (for more information, see Fu et al., 2010). “We hide our identities so that they don’t know we are spotting them,” Chen said. It is also necessary to find ways to disguise the fact that the crawlers are pulling out so much information, as it can cause the web servers running the terrorist sites to slow down. There are various ways to accomplish this, Chen said. “Our collection is about 10 terabytes in size,” he said. “We are close to about 5 billion pages of messages, all generated by [the potential terrorists], not by news reporters.” The content is drawn from about 10,000 websites. Much of the content—about 80 to 90 percent—is political discus- FIGURE 3-1 Dark Web Forum Crawler System. SOURCE: Fu et al. (2010, p. 1,219). Reprinted with permission from John Wiley and Sons, http://interscience.wiley.com [January 2011].
OCR for page 20
Sociocultural Data to Accomplish Department of Defense Missions: Toward a Unified Social Framework - Workshop Summary sion or social discussion, but a small percentage of it is violent and radical. When his group first identifies a new site of interest, they investigate it thoroughly, downloading all the relevant information. After that they update information from the site about once a week. The content they collect includes a large variety of file types, Chen said. Many of them are indexable, with written words that can be recognized by a computer: HTML files, Word files, PDF files, and so on. “What is more interesting are the multimedia files—the images, the audios, and videos.” They make up 40 to 50 percent of the files collected in terms of volume, that is, the number of bytes of storage they require. “These are extremely graphic,” Chen said, and many of them are focused on improvised explosive devices and other weapons. “They show how to create them, they show their qualities, and so on.” As they are collecting these files, Chen said, they are also recording the relevant information about the files: the metadata. “We know who sent it at what time. We know their screen name, we know their font size, we know the font color, we know the PDFs, we know the image file, we know the size—everything.” They then analyze the files in a number of ways, all of it done automatically with computer software. For example, they use different computational linguistic tools to find out the sentiments expressed and the ideas being discussed in the various files. They also look at the textual features, the style, and the genre of the writing. In particular, Chen discussed identifying two types of sentiment: hate and violence. Working from the text of the message, they pull key words from the text and then use them to define a hate score and a violence score. “There is a big paper describing how you define violence and sentiments,” he said (Abbasi et al., 2008). Once the messages have been scored for various emotions, the scores can be analyzed mathematically. As an example, Chen showed the results of a correlation analysis between hate and violence for groups in the Middle East and groups in the United States (see Figure 3-2). In both cases there is a clear correlation between the hate and the violence scores in the messages, Chen said, but the correlation was much stronger for the Middle Eastern groups than for the U.S. groups. Furthermore, there were significantly more violent sentiments expressed by the Middle Eastern groups than the U.S. groups. Another type of analysis creates a profile or signature for each message. Because each writer tends to string words together in slightly different ways, if one has enough writing from a given author, it is possible to determine a set of characteristics that sets that writer apart from others and enables the identification of other messages written by the same writer. “For English text you can reach about 95 percent accuracy, for
OCR for page 21
Sociocultural Data to Accomplish Department of Defense Missions: Toward a Unified Social Framework - Workshop Summary FIGURE 3-2 Measuring hate and violence: United States versus Middle Eastern groups. SOURCE: Chen, unpublished data, Eller College of Management, University of Arizona. Arabic about 92 percent,” Chen said, but it is necessary to have 5 to 10 messages by the same writer in order to create the signature. It is particularly useful to analyze the social networks revealed by the various postings and responses. From the interactions among the various actors, it is possible to determine such factors as how central or how important a person is to the network of other people with whom he or she interacts. This in turn can be combined with such measures as how violent a particular person’s messages are. The resulting analysis has shown, for example, that the violence level of a user is very stable and is very hard to influence by other users. However, Chen said, as people spend more time in the Dark Web forum, their language tends to become more violent. “Violent users try to incite violent ideas by starting many threads,” he said, “but typically other people don’t participate. So they reply to as many threads as possible to try to introduce their ideas into other peoples’ threads.” Based on this work, Chen has created a web-based data set and set of tools that can be used by people interested in learning about terrorists. The current version of this Dark Web Forum Portal contains 13 million
OCR for page 22
Sociocultural Data to Accomplish Department of Defense Missions: Toward a Unified Social Framework - Workshop Summary messages posted in 29 major jihadist forums in Arabic, English, French, German, and Russian. The 29 sites have 339,000 members, but about half of them do not take part in any of the discussions. They seem to be just lurking and listening in on the conversations. The tools provided with the portal allow users to analyze the contents in various ways. There is a search tool, for example, that can be used either inside a single forum or across multiple forums to find threads or individual messages that contain particular key words. Google translator1 will translate key words for searchers in various languages or translate messages and threads from one language to another. Network analysis tools make it possible to analyze and visualize the connections between individuals as revealed by their interactions on the forums. It also identifies particularly important individuals as indicated by the number and types of interactions they have with others. A sentiment analyzer tool performs sentiment and affect analysis of forums, measuring member opinions and emotions. And a text analyzer looks at various language characteristics in messages, displaying the results using various visualization tools. In response to a question, Chen explained that he validates his tools with standard computer science methods. “You reserve a set [of characteristics] as the gold standard that you know the right person’s identity, and you keep it separate,” he said. “Typically, you look at performance metrics of accuracy, recall, and precision.” A better question, he said, is whether the techniques are actually useful for a given problem. “That is actually a lot more difficult question.” Answering it requires testing his techniques against other techniques on such measures as how quickly a person can be identified. “System time savings, case closures—those are the things that people in the field care about.” And for some of these tools, he said, he has indeed shown that they are useful methods.2 Finally, Chen briefly described a similar project he has begun that looks at political, social, and economic issues rather than jihadist ones. At the moment, this Geopolitical Web is focused only on issues relating to the countries of Yemen and Somalia. “I wasn’t brave enough to look at Iraq and Afghanistan,” Chen said dryly. “The framework is very similar [to the Dark Web]—collecting economic indicators, political indicators, and country indicators,” he explained, “but I started with mass media and news. I also collected 1 See http://translate.google.com/ [October 2010]. 2 The set of tools that Chen developed to explore the Dark Web is available for registered users through the University of Arizona website. See http://ai.arizona.edu/research/terror/ [October 2010].
OCR for page 23
Sociocultural Data to Accomplish Department of Defense Missions: Toward a Unified Social Framework - Workshop Summary chatters in the social media, but these chatters are different from Dark Web.” While most of the Dark Web information comes from forums and YouTube, the data for the Geopolitical Web is taken from forums, blogs, and Twitter messages. “Twitter is very interesting. These are other things that [users posting on Twitter] care about at a particular point in time.” As with Dark Web, Chen uses various tools to characterize and represent the data in the Geopolitical Web and then to analyze it. “We have topic, sentiment, time-series data, and social data,” he said, “and then we use various metrics to do geopolitical risk identification.” One goal of the project, he said, is to see whether he can use all these data to predict risks of various sorts—economic risks, political risks, risks of terrorism, and so on. It is potentially a very valuable tool, he said, but he and his group need help from collaborators of various types—political scientists, economists, military and intelligence analysts, among others—in order to take full advantage of it. “We are mostly computer scientists,” he said. They would like to collaborate with others who can help them take this work to the next level. IDENTIFYING CORE OFFENDERS According to David Kennedy, director of the Center for Crime Prevention and Control at John Jay College, in the first couple of years of the 21st century, the murder rates in Chicago’s most dangerous neighborhoods were some of the highest in the country. By 2004, however, they had dropped to about a third of their previous level, not much higher than the rates in some of the safest parts of the city. It is a striking example of how sociocultural knowledge and techniques can be applied to identify the most serious offenders in an area and control their most dangerous behaviors, and it may well have lessons that can be applied in military conflict environments. Kennedy explained that the dramatic decrease in homicide rates was the work of a group led by Tracey Meares at the University of Chicago—and it was no accident. The same techniques have been used to sharply decrease homicide rates in Boston and other cities as well (for more information on the Boston Gun Project, see Kennedy, 2002). “When I got into this field 25 years ago, it was the very firm belief both of people in criminology and people in criminal justice that law enforcement couldn’t do anything about crime,” Kennedy said, “and they were right at the time. All the research and all the field experience showed that. That is not true anymore.” Today, there are two principal approaches to preventing crime that have been widely implemented with proven effectiveness. One involves
OCR for page 24
Sociocultural Data to Accomplish Department of Defense Missions: Toward a Unified Social Framework - Workshop Summary what looks like ordinary police work, but it is motivated by and carried out with stringent attention to up-to-date information. “When I started in this work,” he said, “police departments would get their crime data nine months after the crime happened. Now they get it hourly, and they move people and resources and existing capacities around accordingly, and it turns out that that works very consistently.” The second approach that has been shown to work is to focus on a particular problem in great detail and, informed by theory and research and data, to find ways to address it, always being willing to try new approaches or abandon old ones as necessary. This is the approach that has been successful in bringing down homicide rates in a number of cities. In developing this approach, Kennedy said, he did not start with some overarching theory or unifying social framework. “It began with on-the-ground observations of what was happening to get people killed, and it drew on enormously rich insights of front-line people in law enforcement and communities, which led to a particular picture of what was happening.” This picture of what was happening did not exist in any of the social science or criminological literature at that point, he said, and “if we had started with those literatures, we never would have gotten anywhere.” However, once he and his team had a good idea of what was going on, they found that a great deal of existing research in various disciplines was helpful. The problem that Kennedy focused on was very specific. “The beginning of this was in the crack epidemic,” he said. “There were bodies falling right and left. This was about stopping that kind of serious violence.” In short, he set out to reduce the number of killings in the most dangerous parts of town. He was not trying to do something about gangs, to fix neighborhoods, or to deal with poverty or racism or other broad issues. “Those are good things to do,” he said, and “they are logically connected with these bad outcomes.” But history shows that efforts to lower the homicide rate by dealing with these other issues have uniformly failed. The first step in tackling the problem was to understand exactly what was going on, and so Kennedy analyzed the crime data and spoke extensively with both the people living in the neighborhoods and the police who worked in them. “Probably the most important thing that this on-the-ground work discovered was that this is all about groups,” he said. For example, in Cincinnati he has identified 60 groups of offenders operating in the city who are the focus of his program there. “These 1,500 people are associated as victims, offenders, or both with 75 percent of all killings in Cincinnati,” even though they represent only 0.3 percent of the population. Thus it is important to learn exactly who these groups are and how they operate. One of the most important things to understand about these groups is that, although they are generally referred to as gangs, their structure and
OCR for page 25
Sociocultural Data to Accomplish Department of Defense Missions: Toward a Unified Social Framework - Workshop Summary behavior are not what most people associate with gangs. “They are largely leaderless, they are largely inchoate,” Kennedy said. “Most of the violence is about personal back-and-forth stuff. It is not in the interest of the group; it actually causes damage to the group. There is nobody in charge, and when there are people in charge, they don’t actually control very much. I had a guy who was Mexican Mafia in prison say to me, ‘Do we green light things from here inside? Yes, we do. What people say about us is true. Do we control very much? No, most of what goes on out there is just local random street stuff. We couldn’t control it if we wanted to.’” Even if the groups are not gangs in the usual sense, the group dynamics still lead the group members to behave in ways that differ from how they would act as individuals. “Groups matter, because groups are not the same as individuals,” Kennedy explained. Groups behave very differently. In this setting, groups carry the street code. Violence is driven by a street code. It is the groups that define that, carry it, and impose it on their members. The groups carry the vendettas and the rivalries that generate a lot of the violence.” Social psychology research has shown that groups tend to extreme behavior—that is, people in groups tend to behave in ways that are more extreme than they would behave if they were by themselves. In theory, Kennedy said, groups may tend to become either quieter or more agitated, but in practice the tendency is toward greater agitation. So people who are already more likely than others to get into trouble, or become violent, become even more so when they are in a group. Another characteristic of groups is a tendency toward pluralistic ignorance, or a situation in which everybody in the group believes that everybody else in the group believes something that nobody in the group actually believes. “These guys say all the time, ‘I don’t care about going to prison,’ and we take them seriously. No sane human being wants to go to prison, but they all think their boys believe it.” Everyone in the group believes that everyone else in the group isn’t worried about going to prison, and even though everyone actually is, they all have to act as if they aren’t worried about prison. Studies of these groups have revealed some key facts about their behavior, Kennedy said. First, very little of the violence is about money. People have generally assumed that the violence is driven by the money arising from the drug trade, but that is not the case. “It is vendetta, it is beef, it is respect, it is honor code, it is all this stuff. If it were about money, they wouldn’t do it, because it draws heat. But as long as we don’t understand that, we keep on focusing on these market issues.” There is also a street code that governs much of the behavior of the groups. It is not spelled out anywhere, but the group members know what it is. “It is about disrespect and how you have to respond to it. It is, The
OCR for page 26
Sociocultural Data to Accomplish Department of Defense Missions: Toward a Unified Social Framework - Workshop Summary enemy of my friend is my enemy. It is, We are victims and we are justified in what we do—the power structure has left us no choice. It is, You don’t go anywhere for help, you handle it yourself.” Only a very small number of the group members are responsible for most of the killings, Kennedy said. About 5 percent of the 18- to 25-year-old young men in the bad neighborhoods are in the groups, and only about 5 percent of the group members are behind most of the violence. One implication is that it is impossible to identify the core offenders with some sort of top-down approach that looks at risk factors. “There is a huge population of at-risk kids in those neighborhoods,” he said. “All the boys are at risk. If you try to identify them by identifying risk factors, you get everybody.” So it is important to identify the specific groups and the specific individuals who are most likely to be involved in violence, and the best way to do that is to speak directly with the front-line police officers familiar with the areas of interest. “The formal records and criminal justice agencies collect vast amounts of information,” Kennedy said. “It is just not as helpful in this setting as local people who know what is going on and are willing to share what they have got.” Speaking to the higher ranking police officers is nowhere near as helpful as talking with the cops on the streets, who generally know exactly who the core offenders are. As an example, Kennedy showed a network diagram of the various drug crews in Cincinnati (see Figure 3-3). Each dot is an identifiable group. A straight line between two dots means the groups are allied. A short dashed line means they are feuding. A long dashed line means it could go either way. “This was put together in half a day by sitting down front-line police officers and asking them what they knew about their turf,” he said. “This could not have been produced from official records. You can immerse yourself in formal data and never get here, whereas people close to the front lines, they know all this stuff, and nobody ever asks them.”3 The people in the various nonprofit groups serving these communities have the same information, Kennedy said, and they could serve a similar role. Suggesting that something similar could be done in Afghanistan to identify members of the Taliban, he said that Gretchen Peters, author of Seeds of Terror (2009), claims that “everybody in every Afghan village knows who the Taliban are, and there aren’t very many of them. Could you then figure out from among those who the impact players are? My guess is that you could.” 3 For more information on partnerships between academics and police in Cincinnati to reduce violence, see Engel and Whalen (2010).
OCR for page 27
Sociocultural Data to Accomplish Department of Defense Missions: Toward a Unified Social Framework - Workshop Summary FIGURE 3-3 Network analysis of street sets in Cincinnati. SOURCE: Engel (2010). Reprinted with permission.
OCR for page 28
Sociocultural Data to Accomplish Department of Defense Missions: Toward a Unified Social Framework - Workshop Summary Identifying the groups and their members is the key to getting these groups to change their behavior, Kennedy said, but the focus has to be on the behavior of the groups and not the behavior of the individuals in the groups. The standard law enforcement approach is to catch the individuals responsible for the shootings and send them to prison, but that doesn’t solve anything. “Each of them is a member of a rival set of 20 people. When they go to prison, there are still 19 of their boys on either side. They are still beefing, they are still shooting at each other. You have changed nothing.” The same thing is true for trying to change individual behavior with social services. “You get somebody to renounce violence. He goes back to his neighborhood. The guys who were trying to shoot him are still trying to shoot him, and his boys say, ‘You are going to step up, or we won’t have your back anymore.’ It is not about the individuals.” So how can the behavior of the groups be changed? First, Kennedy said, it is important to have a clearly defined behavioral goal. In this case, the goal is to get members of the groups to stop killing people. It is not to get them to stop dealing drugs or stealing or fighting. Kennedy described the approach he has developed as a partnership of law enforcement, community figures with standing in the eyes of the offenders, and the helping professions. The partnership opens a sustained formal relationship with the groups of interest, with plenty of face-to-face interaction in which the groups are told that the shooting and the killing must stop. Kennedy observed: “My mother says she invented this, and she is basically right. That partnership says to those groups, ‘What you are doing is wrong. We expect better of you. We would like to help you. We are not asking—this is not going to happen anymore.’ Then you stick with it.” The engagement with the core offenders has three prongs, he went on. First, there must be consequences. They are presented as a last resort, but they are also presented as inescapable: “If nothing else we are doing works, you are not going to get away with this.” The sanctions are explained ahead of time, and it is made clear to the members of the group that any legal means will be used to impose sanctions that they won’t like. Because it is the group’s behavior that must change, the sanctions should be aimed at the group, Kennedy said, and they must be sanctions that matter to the group. “The most fundamental thing about deterrence is that the only thing that matters is what matters in the eye of the recipient. If you say to these guys, ‘You cross the line, and two years from now you are likely to end up in federal prison,’ they don’t care. If you say, ‘Everybody on probation is going to be on curfew for the next six weeks, we are going to do home visits, and you are not going to be able to run
OCR for page 29
Sociocultural Data to Accomplish Department of Defense Missions: Toward a Unified Social Framework - Workshop Summary with your girlfriends and smoke dope with your boys,’ they go ballistic. You have got to look at it from where they are.” The second prong is explicit moral engagement. “This is the last thing anybody thinks you can do with these sorts of people,” Kennedy said. “The cops think they are psychopaths, and you don’t talk with psychopaths. The helping people think they are victims, and you don’t blame victims. Nobody ever says to them, ‘You shouldn’t shoot people.’ Nobody says it, so they think it is okay. You treat them like rational people. You say the simple mother stuff: ‘This is not okay. We care about you. We reject a couple of things you are doing. The ideas you are living by are stupid and wrong and self-destructive.’” This is most effective, he said, if the message is delivered by people whom the group members already care about: mothers, ministers, activists, and so on. The reason that these people are not already delivering that message, Kennedy said, is because they do not like the police and believe the police are corrupt and trying to hurt the community. “When strong figures in the community don’t sanction that sort of thing, it is generally because they are so angry at the outside that they are not willing to stand with the outside against their own.” At the same time, the police tend to have their own distrust of and wrong ideas about the community. So it is necessary first to address this mutual distrust and dislike by bringing law enforcement and the local community together to talk explicitly about the misunderstandings and then to turn the talk to common ground and mutual goals. “You can do what is effectively a reconciliation process,” Kennedy said. “Everybody can agree that it is better if neighborhood people say ‘Put your guns down,’ and that works, rather than having the cops sweep them in.” It is particularly effective when a mother stands up and describes how it affected her when her son was killed, “and these guys who everybody has given up on sit there with tears streaming down their faces. They are not out of reach.” The process has clear implications for such places as Afghanistan, Kennedy noted. Just as people in the inner-city neighborhoods of Cincinnati and Boston have beliefs and stories about the police that make them suspicious of their motives and unwilling to cooperate, the Afghan population is leery of American troops. “What I hear from my friends who do work on the ground in Afghanistan,” he said, is that “it is common belief that the American military is there to take over the opium trade and keep the money for themselves and that Islamic women are being kidnapped and prostituted on American military bases…. If you don’t deal with what people think, you are not going to deal with the rest of the structure that gives a few people room to blow up American soldiers. My guess is that if we took these community narratives in
OCR for page 30
Sociocultural Data to Accomplish Department of Defense Missions: Toward a Unified Social Framework - Workshop Summary Afghanistan seriously, and dealt with them, that that might hold hope of changing.” The third prong is to offer help to anyone willing to accept it. Help is generally not very effective in these situations, Kennedy said, but it changes the moral calculation. “Neither the community nor the street guys, if that help is seriously tendered, get to say anymore ‘I am a victim, nobody will help me, so I am justified in what I am doing.’ Even if they don’t get a job, they feel more obligated to put their guns down.” There are a number of lessons learned from these successful core offender programs that could be applied in military conflict situations, such as the current conflict in Afghanistan, Kennedy suggested. First, it is vital to have a specific problem focus. “I don’t have a clue what to do about Islamic fascism, but if what we want to do is stop people blowing up Americans and their peers in public and flying airplanes into buildings, that is something I can imagine doing.” Understanding the problem will require learning about the groups involved, the networks of groups, and the network dynamics. And it is important to see the groups and networks clearly and to deal with them as they are, not as we think they are, Kennedy said. Gretchen Peters has told him that “the Taliban has turned into a drug organization. They pay people to plant IEDs [improvised explosive devices]. We are not dealing with ideology here. We have to just see it for what it is.” Furthermore, it is likely that only a very small number of key people need to be dealt with. Again quoting Gretchen Peters, Kennedy said that there is good evidence that this is the case in Afghanistan. “My understanding is that there are about six Taliban leaders sitting there in Pakistan raking in the money from the heroin trade. If I were in the U.S. government, I would go to them and say, ‘If the bombing doesn’t stop, I will find your money, and I will take it. Thank you for your time.’ Which we do on the street, and it works.” Finally, it is necessary to deal with the ideas in the minds of the people on the other side. Ideas matter and they need to be addressed, Kennedy said. “We do not have to believe them to be correct. We need to take them seriously, and we need to recognize that this is a bilateral relationship. It is not them being wrong—it is the relationship that we have with them and the way we are driving each other to bad places.” IS IT TIME TO ESTABLISH A DISCIPLINE OF SOCIOCULTURAL INTELLIGENCE? In what was perhaps the workshop’s most impassioned presentation, Kerry Patton of Henley-Putnam University argued that the single most important step that can be taken to put sociocultural knowledge to work
OCR for page 31
Sociocultural Data to Accomplish Department of Defense Missions: Toward a Unified Social Framework - Workshop Summary in military missions would be to establish a new discipline of sociocultural intelligence, or SOCINT.4 “Very simply put,” he said, “everything that we are doing right now, every single aspect of coming together and brainstorming different methodologies and different modeling programs and analytical tools to understand different social systems and networks, has been going on forever…. When are we going to stop brainstorming and start implementing?” Thanks to study groups, committee reports, and workshop summaries, there is already plenty of information available on ways to put sociocultural knowledge and techniques to work. The problem, Patton said, is that the knowledge and techniques are not getting to the people who need them. “I have been home now [from Afghanistan] for almost two years, and I have been studying all these reports that different organizations are creating on human and social factors. I have asked my contacts in Afghanistan who are still there, have you read these reports? ‘No.’ ‘Why have you not read them?’ ‘Because they are not getting down to our level.’” The people who need the information are not getting it for a simple reason, Patton said: there is no group of people in the military or the intelligence community who have the responsibility for dealing with this sort of knowledge. “We have HUMINT [Human Intelligence], we have SIGINT [Signals Intelligence], we have MASINT [Measurement and Signature Intelligence], we have all these INTs, and they work…. But we don’t have a SOCINT, a sociocultural intelligence discipline in our national security apparatus.” When he has spoken with people in the military and the intelligence community about whose responsibility it is to pay attention to socio cultural intelligence, Patton said, he has received different answers. “I have been told civil affairs operators, this is part of their responsibility. I have been told foreign area officers, this is part of their responsibility. PSYOPS [psychological operations] folks, it is part of their responsibility.” But none of these people has the training or the time to deal with sociocultural intelligence, he said. “If I went up to a SIGINT-er and I said, ‘Mr. or Mrs. SIGINT-er, I need you to build a school, I need you to build a micro-hydro dam, I need you to build a hospital in a war-torn country, and while you do this, I still need you to collect SIGINT.’ Is that even possible? Obviously the answer is no.” When someone from civil affairs or PSYOPS is asked to collect sociocultural intelligence as well, Patton said, they are so overwhelmed with their primary duties that they skip out on the sociocultural work. There is so much emphasis on collecting intelligence on high-value targets that no one has time to worry about the SOCINT aspect. Thus, although plenty 4 For more information on Patton’s views on SOCINT, see Patton (2010).
OCR for page 32
Sociocultural Data to Accomplish Department of Defense Missions: Toward a Unified Social Framework - Workshop Summary has been done to create a fundamental knowledge base in the sociocultural area, “the people who need it most are not getting it.” If the case for SOCINT is to be made successfully, Patton said, it will require an economic aspect. “When the U.S. government wants to introduce something, you better be able to pull out numbers of how this could help out the country economically.” But, he said, there are plenty of examples from history in which sociocultural intelligence—although not known by that name at the time—more than paid for itself. “How many of you remember the banana wars in South Central America? Was that not one of the greatest economic adventures that we ever undertook during that time period in United States history? Easily it could be argued that it was.” “During the cold war, we did very unique operations throughout all of Africa. Some people didn’t like our operations there, but economically it did pan out. Same thing with South Central America.” The key to these successes was the use of “unconventional human intelligence operators,” Patton said, but after the investigations of the Church and Pike committees during the mid-1970s, such activities were no longer allowed.5 Such work needs to return, he said, but it clearly must be done while obeying the restrictions placed on intelligence activities since the 1970s. And the way to do that, he said, is to create the new field of SOCINT. In response to a question about whether the tools for SOCINT already exist and have been field-evaluated, Patton responded that the history of this approach shows that the tools clearly exist to get the job done. “If you look at the history of organizations that implemented sociocultural intelligence initiatives, you will see they are successful. Those that do not implement SOCINT, they do not succeed.” Referring to Kennedy’s success in reducing homicides in inner-city neighborhoods, Patton said that SOCINT has been shown to accomplish similar things in foreign cultures. “Everything that Dave was able to do with the street gangs in Los Angeles, I was able to do in Northeast Afghanistan. I have physically interacted with Taliban commanders and subcommanders. Within two weeks of being in Jamaica, I infiltrated [drug lord] Dudus Coke, his entire network. When I was in Nepal, I spoke to the number-one Nepalese Maoist [who] was still alive.” In short, he said, it can be done, and it can be done successfully as long as the individuals carrying out the SOCINT have the proper knowledge and training. In closing, Patton commented that the United States has already carried 5 These were Senate and House committees, respectively, investigating intelligence gathering by the Central Intelligence Agency and the Federal Bureau of Investigation.
OCR for page 33
Sociocultural Data to Accomplish Department of Defense Missions: Toward a Unified Social Framework - Workshop Summary out what he called “a very interesting SOCINT operation.” The Troops to Teachers Program, begun in 2001 as part of the No Child Left Behind Act, helps retired military personnel begin a new career as teachers in public schools. It began with the realization that schools were not doing as well as they should have in specific locations around the country. “So with the Department of Education and the Department of Defense,” Patton said, “we came together and analyzed—through obtaining criminal statistics, divorce rate statistics, single-parent statistics, and so forth—a whole bunch of stuff that relates back to the social and cultural norms in geographic areas. What we found was, teachers did not really have an interest to teach at these low-income school districts. However, military members who were getting out would be willing to teach at these locations, and they had a plethora of experiences.”6 After the program was put into place, analysis of the underperforming areas showed that they had been reduced by almost 80 percent in size. The pockets were still there, but they were much smaller, and they had been contained. The key to the success was to identify the relevant human systems and networks, Patton said. The Troops to Teachers Program provided a way to tie the network of retired soldiers in with the school network. “Unofficially that was a SOCINT operation,” he said. Still, despite its various successes, there is still not a clear understanding that SOCINT is a necessity, Patton said. Part of the reason seems to be that the people now doing analysis of social and cultural systems tend to be academics with no real understanding of the places where their knowledge needs to be applied. “You have read a book or you have read literature about the place, but you have never physically been there,” he said. “You have never physically lived in the environment, and you cannot really explain what it feels like to be there, what the texture on the walls is truly like.” Ultimately, he said, the only way that the United States is going to be successful in places like Afghanistan is to have people on the ground who are familiar with current sociocultural thinking and know how to apply it to practical situations. “We need real-world, timely information that is ground truth,” he said, “and the only way to get that is by injecting human factors into human systems and networks. That is the discipline of SOCINT.” 6 More information on the Troops to Teachers Program is available at http://www.proudtoserveagain.com/ [December 2010].
OCR for page 34
Sociocultural Data to Accomplish Department of Defense Missions: Toward a Unified Social Framework - Workshop Summary This page intentionally left blank.