The Project, or How I Came to Uganda
For all of those who still have questions about what exactly we're doing here, let me e'splain something to ju.
Prehistory
Back when I was a lowly, second-semester freshman, I started a research project. Although I did not truly understand the scale of the undertaking, I set off to investigate, categorize, and otherwise describe the myriad qualities and quantities of newspaper coverage of Africa. With the nudging of my overly supportive project director, I slimmed this rather over-blown, make-a-Ph.D.-pee-her-pants project into the tough-but-chewy analysis of one specific place, from one date to another, and from but a single journal. Here are the salients, in convenient, pithy bulleted form:
My project seemed especially relevant in its contemporary context, as well. It was the spring of 2003, right at the beginning of the debate over the relative merits of print and digital content. The Times was trying to negotiate the waters of being a dual-format medium. As an added bonus, journalism itself was being knocked around a bit. Blogs were popping up like weeds, causing outrage, annoyance, or giggles in the press club. And the Jayson Blair scandal did little for the Gray Lady's street cred. Shortly after the Iraqi invasion that spring, questions began to circulate about the credibility of Times reporting in general. All this formed the background of my research.
It is widely held in academic circles that American media has largely ignored Africa. The sources of this blind spot are so numerous that deducing any single source is much like going at the Gordian knot with tweezers. Rather than speculate about our lack of interest in the continent, I chose instead to examine how the Times engaged readers about events there. Looking through the Times’ indexes, I counted and categorized the coverage of Uganda during the most tumultuous period in its post-Colonial history.
The results were interesting, if only because they raised more questions than I could adequately answer. Coverage was scant, except for instances of armed conflict. Throughout the slow downward spiral of the country (1980-1983), little information was published. Once rebels began a fierce bush war in Luweero, coverage began to pick up again. Worse, the reports either oversimplified the conflict (as both newspapers and bloggers are wont to do) or took sides arbitrarily.
“Why?” was a question that a 19-year-old was unable to piece together intelligently. Certainly, the assumption is that American interest in Africa is roughly comparable to that of, say, Brazilian’s interest in Arab camel races. While this is a tragic underestimation of American fascination with and horror of the continent, that seemed to be the attitude of the Times. Since few reporters take up jobs in foreign bureaus, and fewer stay in them for significant lengths of time, the little reporting that occurred was uninformed. And -- an insight at the time, though it would shock nary a five-year-old today -- coverage was largely sensationalist.
There were so many shortcomings in my report. I should have invested some energy in researching the reporters stationed in East Africa during the time, maybe gotten access to their notes and biographies, but I did not. I should have taken the time to carefully review more of the articles themselves, but I did not leave myself time. And I should have compared coverage with other countries and topics, but I did not think of it. It's a project I really need to revisit and update; it showed so much promise, and I was too inexperienced to handle it appropriately. If I can find an electronic copy, I will be sure link to it.
Recent Past
After two years, I decided to revisit the issue of newspapers and African wars. For my undergraduate thesis, I tackled the problem from a different angle, however. Working from a large collection of news articles (not an index) and focusing on the information content, not on tone, this project was more limited in scope, but more rigorous in the approach. Here again, easily digestible bullets!
The war has been brutal and unpopular. At the height of the conflict, over two million people had been displaced from their homes and were housed under appalling and dehumanizing conditions in Internally Displaced Persons (IDP) camps. Although the conflict has raged since 1986, very little is known about it outside academic circles. And information on the war is hard to come by; good information is even rarer. The popular understanding of the war as an insurrection with the aim of establishing an autonomous Acholi state governed by the Ten Commandments is the pernicious myth that clouds any real understanding of the conflict. In fact, Finnstrom demonstrates the ridiculous nature of those claims.4 But in 2006, there were few accessible ways of understanding the conflict, short of digesting reams of Government of Uganda (GoU) and NGO documents or risking life and limb traveling to the north.
Our first attempt to parse the conflict took two parts: 1) try to develop a narrative of the conflict, based on real incidents, statements, and opinions of people actually involved directly in the conflict; and 2) to find a basic sequence of major events that would help inform later studies. These two tasks addressed to fundamental weaknesses in academic studies of the war. First, only a few voices were recognized as contributors to understanding the conflict. The majority of the dialog was controlled by official press sources and a government and military that were made up largely of non-Luo-speaking peoples from the south of Uganda. Second and equally limiting, no comprehensive chronology of the 21-year span of the war exists.
Using the materials collected by my adviser from two English-language newspapers published in Kampala, I aggregated data from individual newspaper reports from January, 2003 to December, 2004.
The results were interesting, if not entirely conclusive. First, according to the government newspaper, over 93% of all civilian casualties during this period were the result of LRA attacks.5 However, it seems unlikely, considering the history of armed conflicts that over 90% of civilian casualties could attributed to just one side or the other, except in those rare cases where certain, inexcusable predilections are prevalent. Second, over the same period, 2.35 civilians were returned for every civilian reported abducted.6 As I discuss, these numbers undermine commonly held assumptions about the tactics of the LRA. Next, engagements between LRA and the Ugandan Peoples Defence Forces (UPDF) were regular and frequent. As I wrote at the time, "the regularity of contact between the two sides is impressive and calls into question the common accounts of the LRA as a loosely organized terrorist organization using hit-and-run tactics on mostly civilian targets."7 Challenging the assertion that the conflict is cyclical and follows seasonal patterns that favor guerrilla tactics against civilians, I noted that "the clearest trend across the two year period was a linear increase in total LRA deaths."8 Finally, I found that as the numbers of civilians that were "rescued" by UPDF decreased, the number of LRA fighters that surrendered, defected, or were captured by UPDF increased. I concluded that these results are clouded by the murky definitions of “civilian,” “collaborator,” and “rebel” as they are used in New Vision accounts. Because these categories are never clearly defined, and because the incidents are rarely described in much detail, there is often some confusion about whether a group of returnees were armed rebels who surrendered on a battlefield or recent, unarmed abductees “rescued” by the UPDF. In many cases, the reader can make no realistic distinction. Furthermore, UPDF reports often intentionally interpret proximity to LRA forces as evidence of collaboration. By this logic, being in the UPDF’s kill zone means an individual is a rebel.9
The most valuable lesson I took from this research, however, was that statistical history is incredibly satisfying, but horribly labor intensive. It took me six months to quantify just two years of data from one main source, with ancillary information from other newspapers. Extrapolating that rate out across all the relevant materials available on the conflict gives us:
just to go through the published material by hand.
An electronic copy (PDF) of my thesis is available here.
Certain Current Events
There had to be a better way. And, of course, there is. Let a computer count up all this information for us.
The only problem with that is that none of this information was in a computer anywhere. I had worked from printed newspaper accounts. And I'm not about to go typing in 10,000 pages of material by hand. We have to scan all of that material. And a back-of-the-envelope estimation suggests we might have between 1 and 2 million pages of material to go through. That's a lot of scanning. Complicating the situation is the fact that most of this material was produced in Uganda, distributed only in Uganda, and is archived, not in the US, but only in Uganda.
So, not only do we have to scan all of this material in, we have to do it in Uganda. And that's why we're here. We've found a research group that has been compiling, cataloging, and binding newspapers for archival purposes since the late 1980's. They have much of the material we want. And we have wrangled access to some private collections that may have much of what that group is missing. Some of the newspapers we are scanning have also expressed interest in our work. Other archives might be opening up to us soon.
So now we're scanning defunct newspapers because they will be the hardest to find anywhere else. So far, we have access back to 1987. (We want information going back to 1986, but this represents a good start.)
We still have to do most of the work by hand, though. I am personally excited that we have original broadsheets to scan; we insisted on having a scanner capable of handling a full A3 document. But that means reading each day's newspaper and scanning the relevant material. In four hours last Friday, we scanned 35 discreet documents. That was a first attempt, and I spent a fair amount of time trying to figure out how to store the files once we scanned them. I hope to do significantly more than that in the future. (I still hope to take a few thousand documents back with us in August.)
[Note: According to our daily reports, we’ve scanned over 350 files and generated close to 2 GB of data since last Friday.]
And this is just the beginning. I want to index everything we scan. Then when I'm done with that, I want to start dumping all kinds of data into this database. Pictures, oral histories, interviews, video, legal proceedings, pulp fiction, propaganda materials -- basically anything I can condense to text and translate into English will go in here. And then we'll have something worth looking at.
Our full research proposal is available here.
So that’s why we’re here. If you still have some questions, or suggestions, shoot me an email. (The address is in the right-hand column of this page.)
-------------------------------------------------------------------
1. Tito Okello’s surname courtesy of my access to early newspapers at the Centre for Basic Research.
2. It seems to be a peculiar issue of definition, but East African civil wars and rebel movements tend to draw lines between armed groups and political struggles that are clearly linked in spirit and often in leadership. For example, the military command of the Lord's Resistance Army defines itself strictly as an armed group fighting for the political objectives of the unarmed Lord's Resistance Movement. Because of this, and also to contest the assertion that the LRA is an apolitical group seeking only local autonomy, Sverker Finnstrom suggests using the LRM/A acronym in the current conflict.
3. The Wikipedia articles on the LRM/A and the war are pathetic. For instance, although the article on the LRM/A describes the breakdown in the talks as the result of "Kony...simply trying to buy time." (cit.) The article ignores the fact that the 1994 Bigombe talks were largely successful, and that, until the government's ultimatum, the peace agreement was considered largely concluded. I don’t want to endorse them by linking from the body, but they are there to be found.
4. Finnstrom, Sverker. Living with Bad Surroundings. Uppsala University Press, 2003.
5. Rogers, William K. A Statistical Evaluation of the War in Northern Uganda, 2003-2004. Honors Thesis. University of South Carolina, 2006, p. 23. (USC catalog link)
6. Ibid., "Table 2: Civilian Returnees and Abductions," p. 25.
7. "On average, the two armed groups met 16.5 times in a month. With only six exceptions, the government and the rebels fought between 10 and 20 battles per month." Ibid., p. 30.
8. Ibid., p. 34.
9. Ibid., p. 40.
Prehistory
Back when I was a lowly, second-semester freshman, I started a research project. Although I did not truly understand the scale of the undertaking, I set off to investigate, categorize, and otherwise describe the myriad qualities and quantities of newspaper coverage of Africa. With the nudging of my overly supportive project director, I slimmed this rather over-blown, make-a-Ph.D.-pee-her-pants project into the tough-but-chewy analysis of one specific place, from one date to another, and from but a single journal. Here are the salients, in convenient, pithy bulleted form:
- Journal: The New York Times
- Dates: 1979-1986
- Country of Interest: Uganda
My project seemed especially relevant in its contemporary context, as well. It was the spring of 2003, right at the beginning of the debate over the relative merits of print and digital content. The Times was trying to negotiate the waters of being a dual-format medium. As an added bonus, journalism itself was being knocked around a bit. Blogs were popping up like weeds, causing outrage, annoyance, or giggles in the press club. And the Jayson Blair scandal did little for the Gray Lady's street cred. Shortly after the Iraqi invasion that spring, questions began to circulate about the credibility of Times reporting in general. All this formed the background of my research.
It is widely held in academic circles that American media has largely ignored Africa. The sources of this blind spot are so numerous that deducing any single source is much like going at the Gordian knot with tweezers. Rather than speculate about our lack of interest in the continent, I chose instead to examine how the Times engaged readers about events there. Looking through the Times’ indexes, I counted and categorized the coverage of Uganda during the most tumultuous period in its post-Colonial history.
The results were interesting, if only because they raised more questions than I could adequately answer. Coverage was scant, except for instances of armed conflict. Throughout the slow downward spiral of the country (1980-1983), little information was published. Once rebels began a fierce bush war in Luweero, coverage began to pick up again. Worse, the reports either oversimplified the conflict (as both newspapers and bloggers are wont to do) or took sides arbitrarily.
“Why?” was a question that a 19-year-old was unable to piece together intelligently. Certainly, the assumption is that American interest in Africa is roughly comparable to that of, say, Brazilian’s interest in Arab camel races. While this is a tragic underestimation of American fascination with and horror of the continent, that seemed to be the attitude of the Times. Since few reporters take up jobs in foreign bureaus, and fewer stay in them for significant lengths of time, the little reporting that occurred was uninformed. And -- an insight at the time, though it would shock nary a five-year-old today -- coverage was largely sensationalist.
There were so many shortcomings in my report. I should have invested some energy in researching the reporters stationed in East Africa during the time, maybe gotten access to their notes and biographies, but I did not. I should have taken the time to carefully review more of the articles themselves, but I did not leave myself time. And I should have compared coverage with other countries and topics, but I did not think of it. It's a project I really need to revisit and update; it showed so much promise, and I was too inexperienced to handle it appropriately. If I can find an electronic copy, I will be sure link to it.
Recent Past
After two years, I decided to revisit the issue of newspapers and African wars. For my undergraduate thesis, I tackled the problem from a different angle, however. Working from a large collection of news articles (not an index) and focusing on the information content, not on tone, this project was more limited in scope, but more rigorous in the approach. Here again, easily digestible bullets!
- Journal: English-language media published in Uganda
- Dates: 2003-2004
- Country of Interest: Northern Uganda
The war has been brutal and unpopular. At the height of the conflict, over two million people had been displaced from their homes and were housed under appalling and dehumanizing conditions in Internally Displaced Persons (IDP) camps. Although the conflict has raged since 1986, very little is known about it outside academic circles. And information on the war is hard to come by; good information is even rarer. The popular understanding of the war as an insurrection with the aim of establishing an autonomous Acholi state governed by the Ten Commandments is the pernicious myth that clouds any real understanding of the conflict. In fact, Finnstrom demonstrates the ridiculous nature of those claims.4 But in 2006, there were few accessible ways of understanding the conflict, short of digesting reams of Government of Uganda (GoU) and NGO documents or risking life and limb traveling to the north.
Our first attempt to parse the conflict took two parts: 1) try to develop a narrative of the conflict, based on real incidents, statements, and opinions of people actually involved directly in the conflict; and 2) to find a basic sequence of major events that would help inform later studies. These two tasks addressed to fundamental weaknesses in academic studies of the war. First, only a few voices were recognized as contributors to understanding the conflict. The majority of the dialog was controlled by official press sources and a government and military that were made up largely of non-Luo-speaking peoples from the south of Uganda. Second and equally limiting, no comprehensive chronology of the 21-year span of the war exists.
Using the materials collected by my adviser from two English-language newspapers published in Kampala, I aggregated data from individual newspaper reports from January, 2003 to December, 2004.
The results were interesting, if not entirely conclusive. First, according to the government newspaper, over 93% of all civilian casualties during this period were the result of LRA attacks.5 However, it seems unlikely, considering the history of armed conflicts that over 90% of civilian casualties could attributed to just one side or the other, except in those rare cases where certain, inexcusable predilections are prevalent. Second, over the same period, 2.35 civilians were returned for every civilian reported abducted.6 As I discuss, these numbers undermine commonly held assumptions about the tactics of the LRA. Next, engagements between LRA and the Ugandan Peoples Defence Forces (UPDF) were regular and frequent. As I wrote at the time, "the regularity of contact between the two sides is impressive and calls into question the common accounts of the LRA as a loosely organized terrorist organization using hit-and-run tactics on mostly civilian targets."7 Challenging the assertion that the conflict is cyclical and follows seasonal patterns that favor guerrilla tactics against civilians, I noted that "the clearest trend across the two year period was a linear increase in total LRA deaths."8 Finally, I found that as the numbers of civilians that were "rescued" by UPDF decreased, the number of LRA fighters that surrendered, defected, or were captured by UPDF increased. I concluded that these results are clouded by the murky definitions of “civilian,” “collaborator,” and “rebel” as they are used in New Vision accounts. Because these categories are never clearly defined, and because the incidents are rarely described in much detail, there is often some confusion about whether a group of returnees were armed rebels who surrendered on a battlefield or recent, unarmed abductees “rescued” by the UPDF. In many cases, the reader can make no realistic distinction. Furthermore, UPDF reports often intentionally interpret proximity to LRA forces as evidence of collaboration. By this logic, being in the UPDF’s kill zone means an individual is a rebel.9
The most valuable lesson I took from this research, however, was that statistical history is incredibly satisfying, but horribly labor intensive. It took me six months to quantify just two years of data from one main source, with ancillary information from other newspapers. Extrapolating that rate out across all the relevant materials available on the conflict gives us:
3 months of real-time / 12 months of published material * 10 journals * (72 months of published material [estimated average length of existence] / journal)
= 1/4 * 10 * 72
= 180 months
= 15 years
= 1/4 * 10 * 72
= 180 months
= 15 years
just to go through the published material by hand.
An electronic copy (PDF) of my thesis is available here.
Certain Current Events
There had to be a better way. And, of course, there is. Let a computer count up all this information for us.
The only problem with that is that none of this information was in a computer anywhere. I had worked from printed newspaper accounts. And I'm not about to go typing in 10,000 pages of material by hand. We have to scan all of that material. And a back-of-the-envelope estimation suggests we might have between 1 and 2 million pages of material to go through. That's a lot of scanning. Complicating the situation is the fact that most of this material was produced in Uganda, distributed only in Uganda, and is archived, not in the US, but only in Uganda.
So, not only do we have to scan all of this material in, we have to do it in Uganda. And that's why we're here. We've found a research group that has been compiling, cataloging, and binding newspapers for archival purposes since the late 1980's. They have much of the material we want. And we have wrangled access to some private collections that may have much of what that group is missing. Some of the newspapers we are scanning have also expressed interest in our work. Other archives might be opening up to us soon.
So now we're scanning defunct newspapers because they will be the hardest to find anywhere else. So far, we have access back to 1987. (We want information going back to 1986, but this represents a good start.)
We still have to do most of the work by hand, though. I am personally excited that we have original broadsheets to scan; we insisted on having a scanner capable of handling a full A3 document. But that means reading each day's newspaper and scanning the relevant material. In four hours last Friday, we scanned 35 discreet documents. That was a first attempt, and I spent a fair amount of time trying to figure out how to store the files once we scanned them. I hope to do significantly more than that in the future. (I still hope to take a few thousand documents back with us in August.)
[Note: According to our daily reports, we’ve scanned over 350 files and generated close to 2 GB of data since last Friday.]
And this is just the beginning. I want to index everything we scan. Then when I'm done with that, I want to start dumping all kinds of data into this database. Pictures, oral histories, interviews, video, legal proceedings, pulp fiction, propaganda materials -- basically anything I can condense to text and translate into English will go in here. And then we'll have something worth looking at.
Our full research proposal is available here.
So that’s why we’re here. If you still have some questions, or suggestions, shoot me an email. (The address is in the right-hand column of this page.)
-------------------------------------------------------------------
1. Tito Okello’s surname courtesy of my access to early newspapers at the Centre for Basic Research.
2. It seems to be a peculiar issue of definition, but East African civil wars and rebel movements tend to draw lines between armed groups and political struggles that are clearly linked in spirit and often in leadership. For example, the military command of the Lord's Resistance Army defines itself strictly as an armed group fighting for the political objectives of the unarmed Lord's Resistance Movement. Because of this, and also to contest the assertion that the LRA is an apolitical group seeking only local autonomy, Sverker Finnstrom suggests using the LRM/A acronym in the current conflict.
3. The Wikipedia articles on the LRM/A and the war are pathetic. For instance, although the article on the LRM/A describes the breakdown in the talks as the result of "Kony...simply trying to buy time." (cit.) The article ignores the fact that the 1994 Bigombe talks were largely successful, and that, until the government's ultimatum, the peace agreement was considered largely concluded. I don’t want to endorse them by linking from the body, but they are there to be found.
4. Finnstrom, Sverker. Living with Bad Surroundings. Uppsala University Press, 2003.
5. Rogers, William K. A Statistical Evaluation of the War in Northern Uganda, 2003-2004. Honors Thesis. University of South Carolina, 2006, p. 23. (USC catalog link)
6. Ibid., "Table 2: Civilian Returnees and Abductions," p. 25.
7. "On average, the two armed groups met 16.5 times in a month. With only six exceptions, the government and the rebels fought between 10 and 20 battles per month." Ibid., p. 30.
8. Ibid., p. 34.
9. Ibid., p. 40.
<< Home