Bureaucracy of Repression:
The Iraqi Government in Its Own Words

Summary and Conclusions

In two separate shipments in May 1992 and August 1993, eighteen tons of official Iraqi state documents captured by Kurdish parties in the March 1991 uprising arrived in the United States for safekeeping and analysis. A Middle East Watch-led team has conducted research on these documents since October 1992; to date, approximately forty percent of the materials has been catalogued and studied.

This is the first report that discusses these documents. In it, we seek to explain the purpose of the research project, and to offer an assessment of the provenance, physical condition, contents and authenticity of the documents that are currently sorted in the United States. Because of the on-going nature of the project, the findings presented here must be seen as preliminary in nature.

We are also publishing, in the Appendix, a sample of thirty-eight documents, both in the original Arabic and in English translation. This is the first time that most of these documents are being made public. Individual documents obtained by western visitors to the Kurdish region have been published in piecemeal fashion in the past two years. But this report constitutes the first attempt at a systematic analysis of a significant portion of the Iraqi state files.

The documents project is part of a wider effort by Middle East Watch to provide evidence that the Anfal campaign by the government of Iraq against its population of rural Kurds in 1988 amounted to genocide. Middle East Watch is currently working toward the bringing of a case of genocide before the International Court of Justice at The Hague; the documents will constitute one important pillar of evidence in such a case. Other evidence will consist of eyewitness testimonies collected by Middle East Watch in northern Iraq in 1992 and 1993, and forensic evidence obtained there by Middle East Watch in collaboration with Physicians for Human Rights. Significantly, we have been able to make some very important matches between documentary and testimonial evidence, one confirming the other, especially in relation to Iraq's repeated use, in 1987-88, of chemical weapons against the Kurdish civilian population.

The majority of files originate in offices of the General Security Directorate, Iraq's internal intelligence agency, or secret police. The holdings also include significant quantities of files from the military Intelligence agency, and from local offices of the ruling Ba'ath Party. The Ba'ath Party's regional headquarters in northern Iraq, the Northern Bureau in Kirkuk, has been responsible for the implementation of the regime's policies vis-a-vis the Kurds. Thus it is the Northern Bureau that had overall authority in the north for the 1988 Anfal campaign.

The files being examined by a Middle East Watch-led team include for the most part memoranda, correspondence, arrest warrants, background information on suspects, official decrees, activity and investigation reports, logbooks, minutes of meetings, membership rosters, lists of names, census forms and salary tables. Through the mechanism of referencing, the documents are linked to one another in a vast and complex administrative web. Despite their variety, they display a remarkable consistency in style. The language is dry and formal, indicating rigid bureaucratic procedures.

Among the findings presented in this report are three key documents, two of which concern the Anfal campaign, while the third offers a "plan of action" against the insurgency in Iraq's southern Marsh areas in 1989. These three documents are included in the Appendix. Generally, the documents we have found during our research constitute small pieces in a large puzzle. In addition to these three key documents, most notable perhaps among our findings is the unequivocal evidence we have been able to accumulate of Iraq's repeated use of chemical weapons against the Kurds. Likewise, we have been able to unearth an impressive documentary record on the incremental Iraqi campaign to raze to the ground all Kurdish villages and the deportation of their populations.

Other documents cover a range of topics, including the Arabization campaign in the Kurdish areas, military operations in the war with Iran, the activities of the pro-government Kurdish militias, the Anfal campaign, and the political and human rights situation during and after the crisis over Kuwait. The documents provide evidence of collective and extra-judicial punishments, summary executions, and illegal reprisals. Concerning the crucial 1987-89 period, Middle East Watch is confident that the evidence is sufficiently strong to prove a case of genocidal intent on the part of the Iraqi government.

In their totality, the documents attest to the existence of a large bureaucracy which, by the nature of the policies that were carried out against the Kurdish population in the 1980's, was a bureaucracy of repression. Through the documents, Iraq's rulers in the Revolutionary Command Council, the Ba'ath Party and the security apparatus speak with great clarity even if their words are filtered through the bureaucratic vernacular of civil servants following a dull routine of inflexible procedures.

In this report, Middle East Watch is presenting a very small sample of the documents that were captured by the Kurds of northern Iraq in the March 1991 uprising against the regime of Saddam Hussein. They include highlights (for example, the three key documents mentioned above), as well as other documents that show the methodology and routine character of a bureaucracy of repression in action.

Although the interest of Middle East Watch in the Iraqi documents has been one of human rights, the content of these documents goes beyond issues of human rights and civil rights. They offer a unique vista on the inner workings of a sophisticated one-party police state; much analysis remains to be done by other researchers on this aspect of the files. The documents are currently being recorded electronically. Middle East Watch hopes to offer public access to the CD-ROM tapes once the work on the genocide case has been completed - hopefully before the end of 1994.

I. Background: The Kurdish Uprising in 1991

On March 6, 1991, Kurdish citizens in the town of Rania rose up in revolt against the central government of President Saddam Hussein, sparking a popular uprising that spread like wildfire throughout the Kurdish regions of northern Iraq. Soon, all the major Kurdish towns, including the important oil center of Kirkuk, had fallen under the control of the local population and Kurdish rebel parties. Within three weeks, though, Iraqi government troops supported by helicopter gunships returned and crushed the revolt, sending over a million civilians in a desperate flight across the mountains to neighboring Iran and Turkey.

In early April, the allies in the war with Iraq during the Kuwait conflict, most prominently the United States, Britain and France, intervened on behalf of the Kurds. They established a "safe haven" for the Kurds in an area of Dohuk Governorate and forced the Iraqi regime to establish a modus vivendi with the Kurdish rebel parties in the other parts of the Kurdish region. This permitted most of the population to come down from the mountains and return to their homes.

An unstable arrangement between the central government and the Kurds, punctured by small uprisings, lasted until the end of October 1991. Then Iraqi troops, unable to assert central government control, withdrew unilaterally from most of the Kurdish areas, excluding the important city of Kirkuk. Since that time, Iraqi Kurdistan has been under the control of the Kurdish rebels and, following elections in May 1992, of a Kurdish regional government.

II. Capture and Transfer of the Iraqi State Files

These new realities on the ground provided outside observers, including human rights monitors, with unprecedented access to northern Iraq.1 Beginning in December 1991, Middle East Watch has carried out extensive field research and, in conjunction with Physicians for Human Rights, exhumations of mass graves in the area over a period of almost two years to investigate Kurdish claims of genocide by the Iraqi regime in the 1988 Anfal campaign. The organization was greatly aided in this effort when, in the spring of 1992, it came in possession of a large consignment of Iraqi state files that had been captured by the Kurds during the March 1991 uprising.

In the first hours of the revolt, Kurds stormed and took control of offices of the Iraqi government and its agencies, including the various intelligence agencies and Ba'ath Party branches, throughout the Kurdish region. Here large caches of official Iraqi state documents fell into their hands. The Kurdish rebel parties succeeded in moving the majority of thesedocuments from the towns into the mountains before Iraqi troops returned to put down the uprising.

A little over a year later, in May 1992, one of the Kurdish parties, the Patriotic Union of Kurdistan (PUK), agreed to send most of the documents that had come into its hands to the United States through a tripartite arrangement with Middle East Watch and the U.S. Senate Foreign Relations Committee. Under the terms of the agreement, the Foreign Relations Committee turned the documents into official records of the U.S. Congress and stored them in facilities of the U.S. National Archives. For its part, Middle East Watch agreed to lead research on the documents for human rights purposes, including the pursuit of a genocide case before the International Court of Justice, or World Court, at The Hague.

In August 1993, a second Kurdish group, the Kurdistan Democratic Party (KDP), entered into a similar arrangement with Middle East Watch and the Senate Foreign Relations Committee, and a large part of its documents were then airlifted to the United States as well, and stored with the PUK-captured documents. At the same time, a small consignment of six boxes of Iraqi state documents was sent to the Archives by the Unity Party of Kurdistan, since defunct. The PUK cache consists of fourteen tons of documents; the KDP cache: 4 1/4 tons. The total number of pages has been estimated at about four million.

III. Purpose of the "Iraq Documents Project"

Since October 1992, a Middle East Watch team has led research on the Iraqi state files, which cover the period from the 1960s to the summer of 1991, with a heavy emphasis on the 1980s. As of December 1993, some forty percent of the documents had been read and catalogued. The main purpose of this project has been to determine whether the documents give evidence of a clear intent on the part of the Government of Iraq, in the 1988 Anfal campaign, to destroy a section of the Kurdish population because they were Kurds -- in keeping with the language of the 1951 Genocide Convention.2 Middle East Watch is currently working on the preparation of a genocide case against the government of Iraq before the World Court, to be brought hopefully by another government or consortium of governments.3

In its research on the Iraqi documents, Middle East Watch has found hundreds of documents which, taken together, provide the administrative outlines of the Iraqi government's 1987-89 program to identify rural Kurds as a population to be subjected to increasingly severe penalties and, eventually, to be eliminated. The files reveal very clearly the unprecedented concentration of power in Ali Hassan al-Majid, a first cousin of Saddam Hussein and the current Minister of Defense, who, in his capacity as secretary-general of the Ba'ath Party's Northern Bureau in Kirkuk, became the architect of the Anfal campaign.4 To date, no single master plan to exterminate the Kurds has, however, emerged in the collection.

In addition to evidence of intent, Middle East Watch has searched for proof in the documents that genocide was in fact committed by the Iraqi government in 1988. Documentary evidence has been supplemented with testimonial and forensic findings collected in northern Iraq by teams from Middle East Watch and Physicians for Human Rights in 1991-93. Significantly, we have been able to make some very important matches between documentary and testimonial evidence, one confirming the other, especially in relation to Iraq's repeated use, in 1987-88, of chemical weapons against the Kurdish civilian population.

Finally, the team led by Middle East Watch has trawled through the Iraqi state files for evidence of other violations of basic human rights. These include extra-judicial punishments, reprisals against the families of suspected Kurdish guerrillas, torture in detention, the use of chemical weapons, large-scale destruction of homes and property, and forced relocation.

IV. The Documents' Provenance

Virtually all the documents stored in facilities of the U.S. National Archives were captured in the three northern Kurdish governorates of Iraq: Dohuk, Erbil and Suleimaniyeh, and then mostly from the three main towns (bearing the same names) in these governorates. There is also a significant cache of files that came from the town of Shaqlawa in Erbil governorate, where they had been captured from offices of Iraq's internal security apparatus, the Amn, by members of the PUK. Regrettably, the Kurds succeeded in removing only a small volume of documents from the town of Kirkuk in al-Ta'mim governorate, seat of the Ba'ath Party's all-powerful Northern Bureau, before Iraqi troops returned to suppress the uprising there a mere one week after it had erupted. Finally, the holdings in the United States include a smattering of documents from the towns of Tuz Khurmatu (Salah al-Din governorate), Khanaqin (Diyala governorate), and Sheikhan and Aqra (Nineveh governorate).

The vast majority of files in Kurdish hands originate in offices of the General Security Directorate (Mudiriyat al-Amn al-Ameh) in the towns of Dohuk, Erbil, Shaqlawa and Suleimaniyeh. The Amn is Iraq's internal intelligence agency, the secret police, which fell under the Ministry of the Interior until 1989, and from then on reported directly to the Office of the President. The Amn has its headquarters in Baghdad, from which it guides the work of the main Amnbranches in each governorate. The Kurdish Autonomous Region (established by government fiat in 1974), which comprises the governorates of Suleimaniyeh, Erbil and Dohuk and is based in the town of Erbil, has a special Amn office (referred to as the Security Directorate of the Autonomous Region) that reports directly to Amn headquarters in Baghdad on Kurdish matters.5

The holdings also include significant quantities of files from the General Directorate of Military Intelligence (Mudiriyat al-Istikhbarat al-Askariyeh al-Ameh). The Istikhbarat falls under the direct authority of the Office of the President in Baghdad and asserts its authority in the nation through regional headquarters (manthumat). Two of these affect the Kurds: the Northern Sector Istikhbarat, which is based in the town of Erbil and covers Nineveh (Mosul), Dohuk and Erbil governorates; and the Eastern Sector Istikhbarat, which is based in Kirkuk and covers al-Ta'mim, Salah al-Din, Suleimaniyeh and Diyala governorates.

In the labyrinth of Iraqi intelligence agencies, there does not seem to be a clear division of labor; some agencies appear even to have been set up specifically to spy on the activities of the others. Generally, though, it can be said that the Istikhbarat deals exclusively with military matters, while the Amn focuses on the civilian domain. Thus, during the regime's counter-insurgency campaign against the Kurds in the 1980s, the Istikhbarat was responsible for gathering intelligence of a military nature, involving activities by and against the armed guerrillas who were based in the countryside or in neighboring Iran. If armed men were captured by the army during military operations, they would invariably be handed over to the Istikhbarat for questioning. It seems, though, that even civilians arrested by the army would end up with the Istikhbarat, not the Amn. By contrast, the Amn would operate in the towns and hunt down civilian members of the Kurdish parties, referred to in the official language as the "internal organization" (al-tanzim al-dakheli), or urban underground. The Amn might also accompany army units during actions against villages because of its specialized knowledge and expertise, and persons captured during such missions might end up in the hands of either the Amn or the Istikhbarat.

The Iraqi state files also comprise documents from local offices of the Ba`ath Socialist Arab Party (Hezb al-Ba`ath al-`Arabi al-Ishtiraki).6 Technically, these are not government files, since the Ba`ath Party is independent from the state and its structure remains strictly separate from that of state institutions. But Iraq has been ruled by the Ba`ath Party since 1968, and for all practical purposes the Ba`ath Party, through its secretary-general, Saddam Hussein, has the final word on all major issues affecting the country. The Party has a mass membership, extending throughout public institutions, the armed forces, places of work, educational institutions and local communities. Party membership is an Iraqi citizen's ticket to job promotion, but the reverse is also true: once special efforts are made to recruit a particular person, refusal to join may trigger the loss of that person's job. In higher education, thesituation was worse in the 1980s: to continue one's studies beyond a Bachelor's degree was impossible if the student was not also a Ba`ath Party member.

The Ba`ath Party maintains its ideological grip on Iraq through its regional bureaus. The one affecting the Kurds is the all-powerful Office of the Organization of the North (Maktab Tanzim al-Shimal, the "Organization" being the Ba`ath Party) based in Kirkuk, here referred to as the Northern Bureau. The function of this office has been to implement and even formulate, the regime's policy in the Kurdish areas, especially during the mid-to-late 1980s. Under its secretary-general in 1987-89, Ali Hassan al-Majid, the Northern Bureau drafted and directed the campaign to crush the Kurdish insurgency, and thus became directly responsible for the murder of tens of thousands of Kurdish civilians.7 The Kurdish parties claim that some of the most incriminating documents and audiotapes currently being analyzed were taken directly from al-Majid's home in Kirkuk. Middle East Watch has seen no files that originate in the office of the Northern Bureau, but we have found a number of original Northern Bureau documents in Amn and Istikhbarat files.8

The cache of Iraqi documents being analyzed by a Middle East Watch-led team also includes a sprinkling of files from the following agencies (or original items of correspondence from these agencies found in files belonging to other agencies):

· -The Presidential Cabinet of the Republic (Diwan Ri'aset al-Jumhuriya), the office of the President, whose executive orders are habitually signed by the office's Secretary (individual items of correspondence only).

· -The Northern Affairs Committee (Lajnet Shu'oun al-Shimal) of the Revolutionary Command Council (Majlis Qiyadet al-Thawra) in Baghdad, which oversees the RCC's policy in the Kurdish areas (individual items of correspondence only).

· -The Command of the Office of the Organization of the North (Qiyadet Maktab Tanzim al-Shimal), a small but powerful office based in Kirkuk that coordinates activities between the RCC's Northern Affairs Committee and the Northern Bureau (individual items of correspondence only).

· -The Central Intelligence Apparatus (Jihaz al-Mukhabarat al-Markeziyeh) in Baghdad, Iraq's foreign intelligence agency, whose northern branch is based in the town of Tikrit, with smaller branches in Suleimaniyeh and other towns (individual items of correspondence only).

· -The National Defense Contingents (Afwaj al-Difa' al-Watani), the pro-government Kurdish militias, based usually in the tribal areas from which they recruit their members (complete files).

· -The National Defense Corps Command (Qiyadet Jahafel al-Difa' al-Watani), the command of the pro-government Kurdish militias, which had five regional headquarters in the north (complete files).

· -The Popular Army (al-Jaysh al-Sha'abi), a militia largely made up of "volunteers," created to support the army by manning guardposts at government facilities, and other tasks (complete files).

· -The commands of regular army divisions, including the First Corps (Faylaq al-Awwal) in Kirkuk and the Fifth Corps (Faylaq al-Khames) in Erbil (complete files).

· -The Command of Oil Protection Forces (Qiyadet Quwat Himayet al-Naft), a special army division charged with the protection of the town of Kirkuk and surrounding oil fields (complete files).

· -The Security Committees (al- Lajnet al-Amniyeh), small but powerful regional committees that combine representatives of the local government, armed forces and security agencies in a single forum to coordinate security policy, and are based in the various governorate and district capitals. They served as executive agencies of the regime in the Kurdish areas during the 1980s; they were abolished after Anfal, and then re-instituted during the crisis over Kuwait (complete files).

· -The Committees to Fight Hostile Activity (Lajnet Mukafehet al-Nashat al-Mu'adi), committees similar in conception to the Security Committees but operating at a local level (complete files).

· -State-controlled "popular organizations," like the Youth Union, Student Union and Women's Union (complete files).

· -Local government offices, such as the local Department of Health (complete files).

· -Kurdish "parties" set up by the regime to create the impression of Kurdish opposition to the guerrilla organizations. They include the Kurdistan Revolutionary Party (Al-Hezb al-Thawri al-Kurdistani) and the Kurdistan Democratic Party (Al-Hezb al-Dimuqrati al-Kurdistani). The latter should not be confused with the KDP currently headed by Mas'oud Barzani (complete files).

Finally, the store of documents also includes files of local Iraqi Police headquarters. More often than not, police documents deal with such mundane matters as common crimes, trafficaccidents, car licensing and the like. The police do not appear to have had any role in the counter-insurgency campaign, either in the countryside or in the urban areas.

In addition to documents written and signed by officials from the above agencies, files found in their offices also often contain photocopies, carbon copies and/or handwritten copies of correspondence and memoranda from officials in other towns. These are of special interest if they include, as they regularly do, decisions from the country's senior leadership. We have found, for example, numerous copies of official decrees that were issued by Iraq's highest legislative authority, the Revolutionary Command Council (Majlis Qiyadet al-Thawra), under the signature of its chairman, Saddam Hussein. We have also found copies of documents from the Special Security Apparatus (Jihaz al-Amn al-Khas), an internal intelligence organization based in Baghdad and run by Saddam Hussein's son Qussay, which spies on the other intelligence organizations by planting agents in their midst; and the National Security Council (Majlis al-Amn al-Qawmi) in Baghdad, an advisory group of security experts chaired by Saddam Hussein.

V. The Documents' Physical Condition

Aside from some small bags of audiocassettes, snapshot photographs, rubber stamps and a pair of handcuffs, as well as a collection of military maps of northern Iraq and three reels of film9, the Iraqi state files consist entirely of written documents. Other materials which the Kurds claim they captured, including further audiocassettes and photographs, as well as videotapes, have also found their way to the West, but do not form part of the cache of material currently stored in facilities of the U.S. National Archives.

The documents being examined by a Middle East Watch-led team include for the most part memoranda, correspondence, arrest warrants, background information on suspects, official decrees, activity and investigation reports, logbooks, minutes of meetings, membership rosters, lists of names, census forms and salary tables. These are kept in either file folders, ring binders, or bound ledgers and logs. Folders are most often held together with the help of shoe strings, and pieces of related correspondence are usually attached to one another with pins (rather than paper clips) and sometimes staples. The majority of the documents are in handwriting. The remainder, especially those written by senior authorities, are typewritten. In many cases, a typewritten version of a piece of correspondence is attached with a pin to the same letter in handwriting, the latter being the draft passed by an official to a secretary for typing and mailing.

The documents' physical condition reflects the circumstances under which they were obtained in March 1991. These tended to vary from town to town. At the start of the mass uprising, Kurdish civilians and guerrillas overran buildings housing government offices and state agencies. In some cases, heavy fighting preceded the take-over, while in others the occupants surrendered without attempting to defend themselves. As a result, some buildings were torched and burned to the ground in the heat of battle, while others sustained little or no damage.

In Suleimaniyeh, in the early days of the uprising, the civilian population engaged heavily armed Amn officers ensconced in their agency's headquarters in a violent battle that left scores dead or wounded. Part of the building sustained fire damage, and as a result some of the documents that survived bear burn marks. Other documents were trampled in the melee, or read and then discarded by civilian looters, and left exposed to the rainy spring weather. Only later did members of the parties attempt to gather up scattered files and transfer them to hastily established offices (and later yet, to strongholds in the mountains). The physical quality of the documents captured in Suleimaniyeh is therefore often dismal: they include crumpled, rolled up and torn sheets of paper, sometimes stuck together and with ink splashed across the page due to prolonged exposure to moisture. By contrast, documents originating in the town of Shaqlawa tend to be immaculate, reflecting the take-over there of government offices by disciplined groups of guerrillas who were clearly under orders to protect the files they found and removed.

In areas where the Kurdish guerrillas were not immediately present, civilians often carried documents home with them. Many Kurds were looking for files about themselves, so as to find out who had provided information about them to the government, and what incriminating information might be in the Amn's possession. They were wont to hold onto their own files, while scattering the others in their possession.

Some of the files in civilian hands were retrieved following calls by the Kurdish leadership to turn them in to party offices. But, even today, significant amounts of documents are suspected to exist in the homes of individual Kurds. Some of these Middle East Watch has been able to inspect in situ. Because of these circumstances, we have no definitive way of ascertaining what the total volume of documents captured from the Iraqi government may have been. Likewise, we cannot confirm that indeed the vast majority of these captured documents was transferred to the United States, even though we have reason to believe that this is so. It is clear that the Kurdish parties have kept at least some of the documents in their possession, usually the ones that refer directly to Iraqi infiltration of their ranks and include the names of Kurds who at one stage or another had acted as agents or informers for the central government. Other files may have stayed behind in northern Iraq as well.

VI. The Bureaucratic Web

The Iraqi documents currently being analyzed by Middle East Watch display a remarkable consistency in style. The language is numbingly dry, the format rigidly formalistic. A standard format would be as follows:

Regarding the memorandum from the office of so-and-so (reference number) of (date), as transmitted to us through a cable, labeled confidential and personal, from the upper command (reference number) on (date), decree number x of (date) has been canceled. Please be informed of this and act accordingly. With regards.

Written thus, the documents bespeak the daily tedium of career civil servants hewing closely to established bureaucratic procedure.

The all-pervasive Iraqi bureaucracy manifests itself in another fashion: through the simple mechanism of referencing, the documents are linked to one another in a vast and complex administrative web. Official decrees are issued from up high and passed down the ranks to the lowliest Amn officers in the various branches via memos and cables. Then reports are generated on the actions that were taken in accordance with the directives, and these reports are sent back up the hierarchy, triggering new memoranda, new instructions, new reports. All decrees are numbered, and so is every piece of written communication. Most memoranda make reference to preceding correspondence and orders issued in years past. In the absence of a computerized index system, this vast network was doubtless managed by diligent civil servants endowed with sharp memories who had to be able to place documents in the right subject files and store these in a logical fashion in local offices. Institutional memory must have been a prized talent among Iraqi bureaucrats!10

In a fashion, the meticulous cross-referencing that is characteristic of the Iraqi documents simply reflects the complexity of daily life in a sophisticated modern state. At another level, though, the mere fact that not a single document stands alone, that every reported action can be related back to an earlier decree, attests to a deliberate strategy on the part of civil servants and other agents of the state to be absolved of any and all personal responsibility for any possible violations of the rights of others committed in the name of the Party, the Revolution, or the Republic. The Amn officer stalking the streets of Salah al-Din, for example, wants his superiors to know that the summary execution he carried out against one of the "saboteurs" had been ordered from up high and, secondly, that he did in fact carry out the order.

A Military Intelligence document of March 1988 offers a telling example of this. It discusses the injuring of two shepherds (one of whom was a member of the pro-government Kurdish militia, colloquially known as the jahsh) by an army unit. Their crime: to be found in the prohibited areas with their herd. The document explains:

The orders issued regarding that subject forbid shepherding or moving about in the areas prohibited for security reasons. The unit then opened fire on them.11

Clearly, as directives stream down in a steady flow, those who carry out the orders revert accountability for their actions back onto the shoulders of central power.

The complexity of the bureaucratic web, the repetition down the ranks of orders from on high, the resulting multiplication of key documents in local offices throughout the Kurdish region, as well as the determination on the part of Iraq's medium and lower-level officials to prove that directives have in fact been carried out -- in short, the completeness and sophistication of the Iraqi archives -- have one important, unintended consequence. Together they emphasize that the documents constitute a credible, authentic expression of the state's actions against the Kurds.

VII. What the Documents Say

Aside from dealing with strictly administrative matters, the Iraqi state files captured by the Kurds paint in great detail Iraqi policy and practice vis-a-vis the Kurds in the 1970s and 1980s. They include research studies, instructions, decrees, arrest orders, execution orders, daily and monthly reports, death certificates, minutes of meetings, and a great amount of correspondence linking the various documents to one another and thus keeping the bureaucracy of repression in motion.

Although Middle East Watch in its study of the documents has focused on the crucial 1987-1989 period -- the tenure of Ali Hassan al-Majid as secretary-general of the Ba'ath Party's Northern Bureau -- we have found strong evidence of flagrant abuses of human rights, as well as the racial animus that was to inform the 1988 Anfal operation, throughout the period covered by the documents. In the view of Middle East Watch, the evidence is sufficiently strong to prove a case of genocidal intent.

Most notable perhaps among our findings is the unequivocal evidence we have been able to accumulate of Iraq's repeated use of chemical weapons against the Kurds. To summarize the evidence: we have found several documents that report on specific air and artillery attacks carried out by Iraqi forces with chemical agents against Kurdish villages in 1987 and 1988. These documents match in precise detail testimonial and forensic evidence collected by Middle East Watch in northern Iraq in 1992. The documents are crystal clear, for example, on the issue of culpability for the chemical attack on Halabja on March 16, 1988, in which some 5,000 Kurdish civilians were killed. While some writers in the United States continue mystifyingly toinsist that the attack was carried out by both Iraqi and Iranian forces,12 the Iraqi state documents, which report widely on Iranian military actions, make no reference to an Iranian gas attack on Halabja at all. Instead, they refer to the occupation of Halabja by Iranian troops and rebel forces belonging to the Patriotic Union of Kurdistan, as well as to the subsequent "Iraqi chemical attack on Halabja."13 In one very explicit case, an Istikhbarat document states that "as a result of the bombing by our planes and our artillery on the area of Halabja and Khurmal, approximately 2,000 enemy forces of the Persians and agents of Iran were killed."14

Although we have found many references to chemical attacks, most have not been explicit. For reasons that still remain unclear, Iraqi bureaucrats most often refer to chemical attacks either indirectly, by reporting that Kurdish sources have accused the Iraqi government of having carried out a chemical attack,15 or euphemistically, by referring to Iraqi "special attacks" (hujoum al-khaass) or attacks with "special ammunition" (`etaad al-khaass).16 There are two ways in which Middle East Watch has been able to establish that "special attacks" are in fact chemical attacks. First, we have been able to match the documentary evidence of specific "special attacks" with our testimonial evidence of particular chemical attacks. The matches are numerous and unambiguous. In one document of early April 1988, for example, the Istikhbarat speaks ofthe recent attack on Halabja with "special ammunition."17 Second, some of the documents themselves establish the link between special attacks and the use of chemical agents, by stating, for example, that the KDP obtained 500 gas masks as a precaution against "special attacks."18

Middle East Watch has also been able to unearth an impressive documentary record on the incremental Iraqi campaign to raze to the ground most Kurdish villages - and even a number of towns. This campaign began in earnest with the border clearances in 1977-78, then was extended to all areas under government control in 1987, and culminated with the elimination of most remaining villages in 1988. Their population was forcibly resettled in government-controlled housing complexes, or deported to southern Iraq for periods of time, or, in the notorious Anfal operation, "disappeared."19

The documents present in great detail the various orders that were issued, often on a village-by-village basis, by the country's political leadership to its military forces, the pro-government Kurdish militia, and the security police charged with carrying out the destruction and relocation. The documents also show how these orders were implemented, with what degree of efficiency, what types of snags occurred during the operation, who participated in it, and if there were any casualties among the local population and Kurdish guerrillas resisting it. There are requests from village leaders to spare their villages, citing special circumstances, such as a long history of collaboration with the regime. And there are indeed documents with instructions from the country's senior officials (the Northern Bureau in Kirkuk) which order the army to make exceptions for villages belonging to certain tribes that had consistently proven their loyalty to the regime by organizing its young men in militias. Middle East Watch has also found scoresof arrest warrants for hundreds of persons who were wanted by the authorities for having left the housing complexes without official permission.

It is important to remember that the Iraqi campaign against the Kurds was in its essence and origin a counter-insurgency campaign, but one that eventually escalated to a level that transcends even the most liberal interpretation of the customary-law principle of military necessity. In the end, it led to the deaths of tens of thousands of non-combatant civilians at the hands of government forces. The documents are full of references to the Kurdish guerrillas, who are referred to as "the saboteurs" (al-mukharrebin), their plans, their meetings, their movements, and the names of their relatives living in areas under government control. They also include actions undertaken against them by army forces and the secret police, their arrest and execution, the deportation of their first-degree relatives and the demolition of their homes.20 The files contain interrogation records, court orders, and death certificates provided by local hospitals. They include petitions by families to be permitted to return from the Arab regions of southern Iraq, where they had been transported in mass deportations in previous years, to their areas of birth, as well as letters from persons seeking information from the authorities about the fate of relatives who disappeared following an army sweep of their areas.

There are documents that speak of punishments to be meted out to those who dared to change their officially-registered ethnicity from Arab to Kurd. And there are other documents that list the means by which Arabs should be enticed to move to the predominantly Kurdish city of Kirkuk, whose surrounding district contains Iraq's most significant deposits of oil. There are policy statements concerning Iraq's small Turkoman population, and also concerning the Yazidis, a non-Muslim sect who consider themselves Kurds (and are considered by Muslim Kurds to be Kurds) but have been designated by the regime as Arabs.21 There are also documents about the Eastern Orthodox branch of Assyrians and Chaldeans, both ethnic groups adhering to the Christian religion who have similarly been defined as Arabs by the regime.

The documents show how policy toward minorities evolved over time: One Ba`ath Party file on a Chaldean soldier in the Iraqi army includes documents from 1982, in which the man is said to be an ethnic Chaldean; from 1985, in which he is said to be "Arab/Chaldean"; and 1990, by which time he was referred to as Arab.22

The files contain ample references to the jahsh forces, the corruption that prevailed in their ranks, the Kurdish benefits they received, the duties they often sought to dodge, and despite all this, their utility as a repository for young Kurds unwilling to give their lives for their country in the war with Iran but prepared, as a lesser evil, to work with the authorities to counter the growth of the guerrilla insurgency in the countryside.

The files bring us up to the Iraqi invasion of Kuwait in August 1990, the second war in the Persian Gulf and subsequent uprisings; but these files are far and few between.23 Fewer still are documents that speak of policies in other parts of the country. Nevertheless, Middle East Watch has so far found one significant document from 1989 that outlines the government's policy in the southern marshes. Called the "Plan of Action for the Marshes," it approves, inter alia, a campaign of the poisoning of the water, the burning of homes and the imposition of an economic blockade in the area.24

What is notably missing from the Iraqi state files being analyzed by Middle East Watch is any direct reference to either torture and rape in detention, or the fate of the "disappeared." This means that such documents either do not exist or were considered so highly classified that they were never distributed to the branches in the north but kept under lock and key in central headquarters.25 Middle East Watch is forced to presume the latter; it is not unlikely that thestrongest evidence of genocide will only be found in the event of a change of government in Baghdad and the opening up of security archives there. Yet, we have found documents that suggest what may have happened with the people who "disappeared," and matched with the eyewitness testimonies collected in the Kurdish areas they provide evidence that the people who have failed to return were indeed killed.26

In addition, the regime has made an effort to disguise some of its cruelest practices either by limiting the circulation of sensitive documents through a "Top Secret" classification; ordering the use of standard euphemisms, both in internal documents and vis-a-vis the public;27 or ordering the suppression of important information. We have offered an example of the first method in footnote 16 above.

Examples of instructions to employ euphemismistic language in internal documents (but likely also in public statements) include a document of the Committee to Fight Hostile Activity in Shaqlawa, which orders substitution of the term "removed villages" (al-qura al-mazaleh) for the term "destroyed villages" (al-qura al-muhaddameh).28 In a clear attempt at occluding the reality about the people who "disappeared" during the Anfal campaign to the public, an Amn document from 1990 makes reference to instructions to replace the phrase: "They were arrested duringthe heroic Anfal operations and remain under arrest," with the phrase: "We don't have any information about their fate."29

We have found several documents that order the suppression of information that could expose secret police methods or might reveal the names of perpetrators of illegal actions. For example, a conference of the Eastern Sector Military Intelligence Directorate held on February 11, 1990 instructs subalterns "not to leak office secrets, especially those relating to the fate of the detainees in your custody."30 An Amn document circulated just prior to the Anfal campaign cites an order from the Northern Affairs Committee of the Revolutionary Command Council:

Do not disseminate the text of the directives issued by senior authorities or mention their provenance. It is sufficient to disseminate the content of the directives with the term: "The following has been decided...," in order to permit the pertinent authorities to take their regular position in issuing directives.31

A note of caution is necessary, though, concerning the credibility of the information contained in the Iraqi state files. Agents of the security police and military intelligence, like their counterparts elsewhere, had a natural interest in exaggerating their accomplishments before their superiors. Thus we have found reports of Iraqi military feats and casualties inflicted on guerrilla forces that have not been substantiated by our research in the field. Likewise, the evaluation by Ba'ath Party officials and Amn agents of the organization, activities and morale of the Kurdish rebels tends to expose real weaknesses in their understanding of their adversaries. For example, the documents often understate the capabilities of the rebel parties in the 1980s. Not a single document should therefore be taken automatically at face value.

By contrast, documents relaying official instructions from senior authorities are of real importance, as they demonstrate an express intent on their part to carry out specific policies. To move from proving intent to proving event, methodologically one must then combine a number of sources, including the documents, the testimonies and forensic evidence collected in the field, and, by comparing the findings, try to reconstruct the events that took place.

VIII. A Narrative Road Map to the Discourse of Repression

The bureaucracy of repression has evolved its own unique discourse, the purpose of which has been to criminalize the regime's opponents and their actions, while justifying and even glorifying the regime's own exploits. This discourse is marked by a number of narrative devices: euphemistic terms are employed to describe obviously illegal acts, while adversaries are assigned a number of epithets that build on Ba'athist political rhetoric, symbols of Arab nationalism, as well as a reinvented cultural and ethnic past. To understand, the true content of the Iraqi state documents, a narrative road map is necessary.

During the Anfal, specific actions to be undertaken against a previously targeted group were not spelled out again, once the original decree has been circulated through the ranks, but given a stock phrasing whose meaning must have been clear to all. Directive SF/4008 of the Northern Bureau, for example, which was issued on June 20, 1987, gave instructions to execute persons aged 15-70 who had been arrested in the prohibited areas; subsequent memoranda, sometimes referring to unspecified Northern Bureau instructions, then simply stated the need to carry out "the necessary measures" (narjou ittikhad ma yalzem).32 Thus, a 1988 letter from the Command of the Oil Protection Forces in Kirkuk to the Security Directorate of al-Ta'mim Governorate states:

We are sending to you the families - their numbers are given below - who surrendered to our forces in the area of Sofi Raza on April 15, 1988. Please take the necessary measures against them according to the directives of the Northern Bureau and acknowledge their arrival.33

As the above document shows, mundane, sanitized bureaucratic language is employed to describe truly horrifying events.

Iraq's war against Iran is referred to as "Qadissiyet Saddam," or Saddam's Qadissiya, a name that evokes the first, victorious battle between the Muslims and Persians in the 7th century. The word "Anfal" (meaning "spoils") refers to an even earlier battle won by the first Muslims, also against "unbelievers," in 624 A.D. From early 1988 on, army desertion is termed an "indecency" under the law. The areas that have fallen under rebel control in the 1980s are called "prohibited areas" (manateq mahdoureh amniyan).34 The deportation of Kurds from their villages to government-controlled housing complexes is referred to as the "amalgamation" (tajmi') and "relocation" (tarhil) of the villages to "modern villages" (qura `asriyeh) or "new cities" (mudun jedideh). Ridding an area of rebels is hailed as "purification" (tathir).

The Kurdish guerrillas, widely known as peshmerga in Kurdish (literally: "those who face death"), are never referred to as such; they are routinely designated as "saboteurs" (mukharrebin). By contrast, members of the government-sponsored Kurdish militias are called "fighters" (muqatelin). Starting in 1987, anyone who refused to move out of the "prohibited areas" to the towns and housing complexes was henceforth also referred to as a "saboteur," whether or not he or she was a combatant. The parties to which the rebels belong, when named at all, receive the prefix "agent" (`amil) as in: "the agent Kurdistan Democratic Party." A small Kurdish party set up by the regime is referred to as "the allied Kurdistan Democratic Party" (halif). More commonly, however, the rebel parties are referred to by different names altogether: the Patriotic Union of Kurdistan is known in the documents as "the agents of Iran" (`umala' Iran) while the original Kurdistan Democratic Party is referred to as "the offspring of treason" (salili al-khiyane).35

Persons, civilians as well as combatants, who voluntarily moved from the "prohibited areas" to areas under government control are referred to in the documents as "returnees to the national ranks" (`a'edin lil-saf al-watani). Before Anfal, this phrase meant that they were pardoned and could move into one of the "modern villages." If, however, a "saboteur" belonging, for example, to "the agents of Iran" did not abide by the regime's summons to "return to the national ranks," he would automatically be branded an outlaw whose first-degree male relatives were subject to detention, and whose mothers and sisters might be "relocated" to "the area of the saboteurs" after they had been stripped of their Iraqi citizenship. Once resident in the "prohibited areas," they all were subject to attack, arrest and summary execution by the regime, a policy that found its apogee in the Anfal campaign.

"Saboteurs" who were defeated in combat, on the other hand, were considered to have "surrendered" (salamu anfusahum). Before Anfal, this usually meant that, unless they benefited from an amnesty, as army deserters (harebin min al-khedmet al-`askariyeh), they would be sent totheir original army units for punishment.36 From 1987 on, though, especially in the period after Ali Hassan al-Majid became Secretary-General of the Northern Bureau, many were killed following their "surrender." During Anfal, the regime's rhetoric could no longer keep up with events: the documents make clear that even those who were said to have "returned to the national ranks" were handed over to the Security Directorate in Kirkuk. In all likelihood, from there they were sent to their deaths.37

IX. The Question of Authenticity

The Iraqi government has on at least three occasions over the past two years publicly challenged the authenticity of the documents currently being analyzed by Middle East Watch, claiming they are forgeries. Middle East Watch rejects these accusations out of hand. For the record, though, it may be useful to rebut the Iraqi government's allegations in detail.

On March 5, 1992, the Iraqi Mission to the United Nations in Geneva wrote to Mr. Max van der Stoel, Special Rapporteur of the United Nations Commission on Human Rights on Iraq, stating in part:

4. With regard to the letters allegedly emanating from security departments at Sulaimaniya, Halabja and elsewhere, we wish to point out that, during the disturbances, hostile bodies succeeded in obtaining stationery bearing the letterhead of those departments, which they used to forge letters ostensibly emanating from official bodies. Investigations have shown that those official letters were not issued by the departments referred to and the information contained therein is wholly fabricated. The manner in which one of these letters was drafted shows that it was written by a person with a poor command of the Arabic language, thereby confirming its spurious nature.38

Then, on February 5, 1993, the Iraqi Mission in Geneva referred to the documents in a note verbale to Mr. Van der Stoel as documents "allegedly issued by Iraqi authorities."39 The note verbale included the official response by the Government of Iraq on the matter of the documents, the text of which reads as follows:

We wish to state that a number of falsified documents has been disclosed by unknown circles with a view to undermining Iraq's reputation, as part of the political and media [war] waged against it. Among those documents are the ones we received in connection with the events which, basically, took place during the Iran-Iraq war up to the July 1988 cease-fire.

It is well known that Iraq's eastern and north-eastern borders were scenes of military operations. It is therefore not possible to verify what went on during that period, especially with regard to activities of the saboteurs which were fully cooperative with the hostile Iranian military forces.

As regarding the documents which were sent to us, with the allegation that they are official documents - which in fact [they] are not - we wish to point out the following:

1. Following the all-out aggressive war which was waged against Iraq on 17 January 1990 [sic], the American, British and French forces occupied vast areas in northern Iraq. By the force of weapons, the invading forces assisted the irresponsible elements and saboteurs in assuming control of the area. This foreign occupation of northern Iraq led to the absence of the State's official bodies. Government departments with all their stores of printing machines, stationery and official stamps bearing signatures of Iraqi officials fell in the hands of the saboteurs and American, British and French forces of occupation. Moreover, many officials, who worked in those departments, fell under the mercy of the saboteurs['] gangs and carried out their orders.

2. All that has facilitated and will facilitate for the foreign powers, and the saboteurs under their command, to carry out large-scale forgery, including what has been so far disclosed of alleged documents, and which may be disclosed in the future.

3. Furthermore, the violent and successive events which were imposed on Iraq, during the Iran-Iraq war or during the aggression perpetrated by the allied forces against Iraq, along with the control by the saboteurs of the northern area, and the preceding riots, have all inflicted damage and loss of most of the official documents in the northern area, rendering the competent Iraqi authorities unable to verify the validity of any information or claims requiring response.

Apart from the charge that they are fabrications, the key point in the Iraqi government's position on the documents is the admission that the events in the Kurdish areas in the spring of 1991 led to "damage and loss of most of the official documents" there. From this statement, two possible conclusions flow. If "loss" is meant to denote that the government "lost" the documents, or "lost control" of them, this would suggest that these same documents might well have survived the events in the north and could conceivably -- and, indeed, would in all likelihood -- be in the possession of someone other than the Government of Iraq. If this is the correct interpretation, then the government's statement that most of its documents were lost would contradict its accusation that the documents the Kurds claim they captured are all forgeries. The government cannot have it both ways: either the documents represent a 4-millionpage forgery on a scale previously unseen in history, or they are indeed the documents which the government has admitted losing. In the latter case, there can be little question as to the authenticity of these documents in their totality.40

If, on the other hand, "loss" is meant to denote that the documents were "irreparably damaged," by fire or other form of destruction, then the question arises by what strange coincidence it would be possible for several tons of stationery to survive the uprising relatively unscathed, at a time when the government's undoubtedly vast inventory of written reports, correspondence, personal files and other official documents somehow perished in its totality.41 Moreover, the Iraqi claim flies in the face of overwhelming testimonial evidence, collected by Middle East Watch, that large holdings of written documents were found in and taken from government offices in all the major towns captured by the Kurds during the March uprising.

The Special Rapporteur, in his report to the forty-ninth session of the Commission on Human Rights in February 1993, pointed at a second contradiction in the various statements made by the Iraqi government on the subject of the documents.42 On February 20, 1992, the representative of the Government of Iraq addressed the Commission on Human Rights, claiming that the documents were forgeries because "paper forms carrying the formal emblems" had fallen into "hands not qualified to use them," and that these documents were "written in very bad Arabic and it seems that they were drafted by people who do not master the language, a fact that confirms counterfeit."43 In his reply, the Special Rapporteur made clear that, again, the Iraqi government could not have it both ways: either the documents were slipshod counterfeit produced by poorly educated Kurds, or they were forgeries made with mirror-image precision by captive government employees working under duress. Middle East Watch and others who have examined the documents are wholly convinced that neither is the case.44

The following features support the argument that the documents are authentic in their totality:

1. The sheer quantity of the holdings transferred to the United States: 18.25 metric tons.

2. The variety and complexity of the material, both in form and content: The files include carefully maintained bound ledgers containing codes or journal entries; thick folders comprising hundreds of chronologically organized pieces of correspondence that are carefully dated and numbered, and refer to previous pieces of correspondence and official orders; personal files that offer detailed life histories of thousands of persons under investigation; salary ledgers with carefully worked out calculations; etc.

3. The nature of the material: The documents provide an extremely detailed view of the nature and scope of Iraqi intelligence operations in the Kurdish region over a period of thirty years (though most of the documents stem from the 1980-1991 interval), but few have self-evident value in human rights terms. At least a third of the documents are entirely administrative in nature, chronicling the workaday world of government employees with their constant demands for vacations, promotions or appointments with senior officials. Some fifty percent of the documents relate to investigations of persons and events, but offer no evidence of crimes. Only a small portion of the documents contain specific orders to undertake actions that would constitute clear violations of international law and human rights, and reports of such actions indeed having been carried out. The vast majority of the documents are completely bereft of utility to anyone interested in seeking legal action or redress against President Saddam Hussein's government, and it would appear to be an act of sheer folly to go about re-creating such an immense store of administrative paperwork merely to indict the regime.

There are other questions of a more logical nature: Why would forgers go to the trouble of manufacturing both handwritten and typewritten drafts of the same document, however innocuous in nature, and even include scribbled notes that must have formed the rudimentary outline for these drafts, if it were only the final, official and signed version that could conceivably be used to bring charges against the regime? Why would they include documents that might potentially be a source of embarrassment to themselves, for example the numerous references to kidnappings for ransom of Kurds suspected of pro-government activity by Kurdish guerrillas? And why would they create large amounts of documents, only to then hold onto some of these themselves, claiming that these are of no significance to human rights organizations but only have internal value for their own movement?45

All these points suggest that a large-scale counterfeit operation would not just have been extremely unlikely, but was indeed completely impossible, especially taking into account the chaotic conditions that prevailed in northern Iraq following the March 1991 uprising. Iraqi forces succeeded in reconquering the area within three weeks, driving the Kurds into the difficult mountainous terrain along the borders with Iran and Turkey. In the wake of government announcements of an amnesty and promises of protection by the allied forces, Kurds began trickling back into the towns, most of which were now firmly under government control, in the late spring. The area remained under government control throughout the summer. Finally, at the end of October 1991, Iraqi forces withdrew voluntarily from most of thearea, leaving the Kurds in charge of the major towns, Suleimaniyeh, Erbil and Dohuk, and their surrounding countryside.

Clearly, no large-scale counterfeit operation could have been staged until this date, if at all. But already in November 1991, western visitors were able to inspect the enormous caches of documents held by the Kurds. Middle East Watch itself inspected huge quantities of PUK-held documents in an abandoned school building in Mawat, near the Iranian border, in February 1992. Stored in grain sacks and ammunition boxes, they were covered with dust and showed no sign of having been disturbed in months.

To prove that any individual document is authentic is a more difficult, but by no means insurmountable, problem. First, the possibility exists to verify the age of the ink and paper used to produce the documents. If a case against the Iraqi government or its officials ever comes to court, such physical authentication would be helpful.

Secondly, if the Kurdish parties themselves wished to indict the Iraqi regime, they would want to prove the regime's intent to commit genocide and crimes against humanity. In the absence of resources to manufacture large amounts of incriminating documents, it would be logical that they would draft a few documents that, by their explicit nature, would promptly be recognized as "smoking guns." These would then be inserted into the captured files so as to make it appear that they had always been there. In fact, the evidence Middle East Watch has been able to collect so far is highly fragmented in nature: it is incomplete and spread out over an enormous number of files. But this evidence, however fragmentary, displays a remarkable consistency, and the tiny bits of evidence in fact turn out to constitute so many small pieces in a gigantic jigsaw puzzle. The task of Middle East Watch has been to assemble this puzzle. After examining some forty percent of the documents, the evidence can already be shown to be so compelling as to dispense with the need for anyone to insert one or more "smoking guns."

The two genuine "smoking guns" already found, both of them orders issued by the Ba'ath Party's Northern Bureau, under the signature of its Secretary-General Ali Hassan al-Majid, in 1987, must in our view be authentic. Middle East Watch has found multiple copies of these orders, sometimes with slight variations, in different boxes of varying provenance. The question this finding raises is: Why would forgers go to the trouble of manufacturing multiple copies of the same order, if a single one would have been enough to do the trick; and how would they be able to distribute these copies through the files in various localities, sandwiching them chronologically between documents of similar but different content, while still maintaining the degree of internal consistency these files invariably exhibit?

Frequent references to one of these key documents, the Northern Bureau's letter no. SF/4008, are made in later documents. Throughout the fall of 1987, in 1988 and even during 1989, the order to kill all inhabitants aged 15-70 caught in the prohibited areas by its blatant nature apparently filled unit commanders with uncertainty and hesitation as to its sweeping nature. The order therefore needed to be restated on several occasions, including in an apparently annoyed manner, by headquarters. The hesitation, the annoyance, and the numerous other nuances that mark the general narrative of the documents bespeak a degree of complexity that would be impossible to falsify.

One should also ask why, if the Kurdish parties indeed produced fake "smoking gun" documents and inserted these into the files, they did not produce clearer proof of the top authority's responsibility. Why do most of the documents originate in Amn and Istikhbarat offices if it is the Ba'ath Party that had final authority in the Anfal campaign? Moreover, some of the most egregious violations -- rape and torture in prison, and the mass killings of non-combatants arrested during the Anfal campaign -- are either left unaddressed in the documents, or are alluded to in indirect and obscure terminology.

Finally, there simply is no physical evidence in the rebel-controlled region of northern Iraq that the Kurds were at any time engaged in a logistical exercise of such monumental proportions. It would have been very difficult to conceal such an enterprise, what with the direct and untrammeled access independent observers had to the area at the time.

There is not a shred of evidence that any one of the documents in the possession of Middle East Watch was falsified, much less all four million of them. Apart from the Government of Iraq, no one has claimed that they were forged. And, to date, Baghdad has presented no evidence that the Kurds have done so.

X. About the Documents Included in the Appendix to This Report

In this report, Middle East Watch is presenting a very small sample of the documents that were captured by the Kurds of northern Iraq in the March 1991 uprising against the regime of Saddam Hussein. They include highlights (for example, the two main "smoking gun" documents), as well as some documents that would be considered of lesser significance for any legal case against the government of Iraq, but are helpful all the same in showing the methodology and routine character of a bureaucracy of repression in action.

The thirty-eight documents in the Appendix below include documents captured by the Patriotic Union of Kurdistan, the Kurdistan Democratic Party, and the Unity Party of Kurdistan. They have been organized in the following categories:

·Arabization: Documents 1-3.

·Policy Toward the Prohibited Areas: Documents 4-5.

·Destruction of Villages and Towns: Documents 6-8.

·Chemical Attacks: Documents 9-12.

·The Administrative Framework of Anfal: Documents 13-19.

·The Anfal Campaign: Documents 20-29.

·The War Over Kuwait and the Subsequent Domestic Uprisings: Documents 30-32.

·Other Documents of Interest: Documents 33-38.

In most cases, documents have been translated in full, but at other times they have been excerpted to reflect the essential information. Editorial comments by Middle East Watch appear in [square brackets]. Some text had to be translated freely to convey the meaning intended in the Iraqi dialect of Arabic when no direct equivalent existed in English. For dates, the Iraqi system has been observed: day/month/year.

Click here to view the Appendix.

1 While outside observers have had unprecedented access to northern Iraq, it has not been without dangers. After the Iraqi government lost control of its northern border crossings with Turkey, Syria and Iran in 1991, it has considered the entry of persons into northern Iraq through these crossings as illegal. Numerous attacks against representatives and facilities of the United Nations and non-governmental organizations are believed to have been the work of agents of the Iraqi government.

2 In July 1993, Middle East Watch published a report presenting the first overview on the 1988 Anfal campaign. This report, Genocide in Iraq: The Anfal Campaign Against the Kurds, combines testimonial and forensic evidence collected during the various field missions to Iraqi Kurdistan in 1991-93 with the documentary evidence that had been garnered as of May 1993 to argue that the Iraqi government did indeed commit genocide against the Kurds in 1987-1989.

3 The need to bring a case before the ICJ goes beyond the important goals of establishing responsibility for the greatest of crimes -- genocide -- which occurred six years ago, and creating a deterrent for other governments that might be tempted to engage in similar crimes. The Kurds might derive practical benefits from a hearing before the World Court as well -- in the form of Court-ordered provisional measures of protection by the international community, Iraqi state reparations to the victims, and hopefully a full accounting by the Government of Iraq for those who disappeared following their arrest in the Anfal campaign.

4 Neither Ali Hassan al-Majid's predecessor as secretary of the Northern Bureau, nor his successor received the broad authority from the Revolutionary Command Council that al-Majid had from March 29, 1987 until April 23, 1989. See Document #13 in the Appendix below.

5 Although by rules of bureaucratic hierarchy the Amn offices of Suleimaniyeh, Erbil and Dohuk governorates were supposed to report to the Amn office of the Autonomous Region (which, as the documents show, they often did), at times they reported directly to the Amn headquarters in Baghdad.

6 "Ba'ath" translates as "Revival." Because of common usage in English, we refer here simply to the "Ba'ath Party."

7 Prior to taking up his post at the Northern Bureau, Ali Hassan al-Majid served as head of the Amn from 1985 to 1987. In the early 1980s, al-Majid was the director of the office of the Regional Command of the Ba'ath Party. As such, he was the third ranking Ba'athi official in Iraq after Saddam Hussein, the secretary-general, and Izzat al-Durri, his deputy. The Regional Command is the leading organ of the Ba'ath Party in Iraq. It forms part of the Arab Ba'ath, which fell under the leadership of founder Michel Aflaq until the latter's death in 1990, when Saddam Hussein replaced him. Following his assignment in Kurdistan, al-Majid served as Minister of the Interior and as Governor of Kuwait during the brief military occupation in 1990-91. Since 1991, he has been Minister of Defense.

8 Kurdish rebels held the town of Kirkuk for barely a week during the uprising. According to rebel sources, they failed to remove most of the documents they had captured in Kirkuk in time because they had not anticipated their quick defeat. Some of the files of the Eastern Sector Istikhbarat were saved, but files from the Northern Bureau apparently were not.

9 The three reels form part of a single film showing a famous Ba'ath Party meeting in 1979, shortly after Saddam Hussein's rise to power, during which a coup plotter-turned-informer tells a large audience the details of a plot to overthrow the new president. Some of the conspirators are present in the audience, and at the mention of their names, Saddam Hussein orders them to leave the room at once. (All, including the informer, are known to have been executed). At one point, following loud exclamations and applause in support of Saddam Hussein, the president, who has been calmly smoking a large cigar, is seen to be weeping. Iraqi exiles say that after 1979, the film was distributed widely among Ba'ath Party offices in Iraq and shown to selected audiences for propaganda purposes.

10 The documents captured in the Kurdish towns show no evidence that a computerized file management system existed in these regional offices. It is very likely, though, that Baghdad headquarters of the various intelligence agencies have been equipped with more sophisticated index systems. The only indication that any sort of organized file management was carried out in the governorates came in the form of a large number of standard-format index cards found in Amn offices. These cards invariably bear a given person's name, date of birth, place of residence, occupation and relevant activity (relevant, that is, to the work of the secret police), and they have a number that turns out to correspond to a file in that same person's name. (Each personal file has the name of one or more persons written on its front cover). Thus the Amn was always able to find a person's file simply by looking up his or her name in the card index. There does not seem to have been a similar system for files of directives, correspondence, reports, or matters of a strictly administrative nature.

11 Letter from the Military Intelligence Directorate, Kalar, to the Military Intelligence Directorate -Eastern Sector (in Kirkuk), ref. #2228, on March 7, 1988. (MEW 2106/4-J).

12 See, for example, Stephen C. Pelletiere, The Iran-Iraq War: Chaos in a Vacuum. New York: Praeger Publishers, 1992.

13 For example, in a telex message from the Military Intelligence Directorate in Suleimaniyeh which refers to a videotape, being sold in several shops in Suleimaniyeh, showing "the Iraqi chemical attack on Halabja." (Ref. 10472 of April 11, 1988; MEW 2107/2-A).

14 In a letter from the Military Intelligence Directorate in Suleimaniyeh of March 27, 1988. (MEW 2106/4-I). "Agents of Iran" is an Iraqi euphemism for guerrillas of the PUK, who had actually captured the town of Halabja, with the help of regular Iranian forces, a few days before the chemical attack. This document is typical of Iraqi documents that describe the campaign against the Kurds in the 1980s in that it conflates the civilian population of Halabja with Iranian troops and rebel forces. On the matter of chemical attacks, this same document also makes reference to an attack in the Qaradagh area in which fifty guerrillas were said to have been killed. We know from our interviews in the field that this was the chemical attack on the village of Sayw Senan on March 22, 1988. (See Document 9 in the Appendix below).

15 Such references are numerous. It usually concerns reports by the Amn of Kurdish accusations expressed before international fora, like the United Nations. The Amn invariably ordered reprisals against the families living in Iraq of those making the accusations. For example, in a letter on June 25, 1987, the Amn directorate of Erbil governorate orders the confiscation of the property of thirteen named Kurds, whom it accuses of having participated in a protest - in front of the Iraqi embassy in London on May 1, 1987 - against "Iraq's use of chemical weapons in the towns and villages in the north." (Ref. Sh.S3/5666; MEW 70/10-F).

16 Although Iraqi bureaucrats clearly tried to avoid making direct references to chemical attacks, our evidence shows that slips did occur. There may have been a general order forbidding them from using the term chemical attacks in direct reference to Iraqi forces. Middle East Watch has so far found one document that orders assignation of the classification "Top Secret" (darajeh `aliyeh min al-ketman) to documents that contain information about the production of chemical weapons. (Memorandum from the Amn directorate in Erbil to Amn sub-directorates, ref. S5/19299 of December 17, 1988; MEW 91/25-A).

17 In a report prepared by the Military Intelligence Directorate - Eastern Sector in Kirkuk, distributed with a cover letter on April 2, 1988. (MEW 2123/5-H).

18 For example, in a document of the Military Intelligence Directorate - Eastern Sector in Kirkuk dated June 19, 1988. (MEW 2128/6-A). It is also worth noting that Ali Hassan al-Majid himself, in one taped meeting in 1987, used the terms chemical attacks and special ammunition interchangeably. (For a partial transcript of his speech, see: Genocide in Iraq, p. 349). One can also draw conclusions about Iraq's use of chemical weapons on the basis of the timing of multiple references to the Kurdish rebels obtaining protective devices.

It is clear from the documents that suddenly, in the spring of 1987, the PUK and KDP began acquiring significant quantities of gas masks and ampules with chemical antidotes. This was not a coincidence: it occurred on the heels of what we know, through testimonies obtained in the field, were the first Iraqi chemical attacks against the Kurds - on the PUK headquarters at Bergalou-Sergalou, near the Iranian border, on April 15, and on the villages of Balisan and Sheikh Wasan one day later. (Ibid., pp. 59-70). Finally, we have found references to Iraqi "air strikes" that, according to the documents, caused people to lose their eyesight. The only logical explanation for a medical complaint of that sort would be the use of chemical agents. (Ibid., p. 71; see also Document 12 in the Appendix below).

19 The three periods mentioned here present three distinct "waves" of village destruction. But indeed villages were destroyed long before 1977 in Ba'ath-run Iraq, and even before the Ba'ath came to power in 1968. Many villages were destroyed in the Penjwin and Chuarta area during the 1980-88 Iran-Iraq war. Even after Anfal, village destruction continued, as the documents clearly show, and the large town of Qala Dizeh was destroyed, along with three adjacent complexes, in the spring of 1989. For detail, see Genocide in Iraq.

20 First-degree relatives include: parents, spouse, and children. It was mostly the women, children and old men who were deported, while the adult men were detained. (See Documents 16 and 20 in the Appendix below).

21 See Documents 1,2 and 3 in the Appendix below.

22 File of Ghazi Shabo Ilia. Ba`ath Party, Einkawa, Erbil (MEW 2188/3-A,B,C).

23 There are a large number of files from the 1990-91 period, but few deal with security matters. We have found a general decline in documents dealing with security matters from 1989 on - after the Anfal campaign had been completed and the regime believed it had solved its Kurdish problem. (See Documents 30, 31 and 32 in the Appendix below).

24 Report distributed by the Amn directorate of Erbil Governorate, ref. S5/1657 on January 30, 1989, and found by Middle East Watch in a box of files originating in Shaqlawa, accompanied by a letter from the Amn directorate there acknowledging its receipt. (MEW 32/1-B). The text of the document has been reproduced in full in Economic and Social Council, Report on the situation of human rights in Iraq, prepared by Mr. Max van der Stoel, Special Rapporteur of the Commission on Human Rights, in accordance with Commission resolution 1992/71 (E/CN.4/1993/45, 19 February 1993), pp. 94-98.

25 We know from testimonies that torture is rampant in Iraqi jails, and that rape in particular may be used as a method of intimidation. It is very likely that these matters are not openly discussed by Iraqi bureaucrats for the simple reason that torture and rape are illegal under Iraqi law. Moreover, the regime's open acknowledgment of such practices, which affect families throughout the country and are not linked to a military campaign against a specific group of people, might stir tremendous unrest among the population.

There has been some discussion in the American media about the documentary evidence of rape in Iraq. The Iraqi writer Kanan Makiya claimed in an article in The Nation on May 10, 1993, that the "General Security Organization" employs persons as professional rapists, offering a single document to substantiate this claim. The document in question is a printed index card with handwritten entries, bearing a man's name, profession and "activity." His profession is said to be "fighter in the Popular Army," and his "activity" is "violation of women's honor."

While there is no reason to doubt this particular card's authenticity, there is no evidence to suggest that the man mentioned on the card was in the employ of the Iraqi secret police at all, for any purpose. Middle East Watch has found hundreds of identical cards in the Iraqi state files. All list the name of a person and his or her profession and "activity." Many of the professions listed are jobs not in the government but in the private sector. The word "activity" is meant to refer to the main reason why the Amn is keeping a file on that person. Almost invariably, the handwritten entry offers one of the following types of "activities": being a "saboteur," membership in one of the rebel parties, fleeing to Iran, "returning to the national ranks," being an army deserter, "arrested in the Anfal operations," or "his father was executed."

The index cards do not constitute a record of employment by the Amn, and in our view neither they, or any other Iraqi government document seen by Middle East Watch, provide a shred of evidenceof the use of rape as a matter of state policy in Iraq. (See also footnote 10 above).

26 We have found multiple documents, for example, about the mass detention of members of the Barzani tribe in 1983, who subsequently "disappeared." Documents from around the same time order employers in Erbil to consider the Barzanis among their employees, who are suddenly absent from their jobs, as dismissed. During Anfal, documents attest to the arrest of thousands of persons and their transfer to Kirkuk. Some of their names match names, given to Middle East Watch, of persons who were said not to have returned home after Anfal, and even include three execution survivors interviewed by Middle East Watch in 1992. (See Document 23 in the Appendix below).

27 See section VIII, "A Narrative Road Map to the Discourse of Repression," for a full discussion of official euphemisms used in the context of the Iraqi government's campaigns in northern Iraq. Some phrasing is meant to conceal unpleasant realities to the public, while at other times euphemistic language is used as a propaganda tool to further the regime's ideological objectives.

28 Letter from the Committee to Fight Hostile Activity in Shaqlawa (ref. LM/135 of May 4, 1988), referring to instructions from the Northern Bureau of April 16, 1987. (MEW 92/1-T). This was at the height of the Anfal campaign, and is therefore likely to reflect an attempt on the part of the regime to conceal from the Kurdish urban population the far-reaching nature of the campaign in the countryside. During operations to relocate villagers to complexes in previous years, the army had often left the houses of villages unharmed, offering the population at least the hope of return. For the duration of Anfal, people were to be left with this illusion, probably to forestall mass protest. Incidentally, a document from 1987, during the spring campaign of village destruction, instructs Military Intelligence agents to "use the term 'amalgamation' for the deportation operation." (MEW 2126/1-S). Again, this change represents a softening of the rhetoric to be used in official documents. One year later, it should be noted, during the much more violent Anfal campaign, even the term "deportation" (or "removal") had become part of acceptable discourse.

29 Memorandum, signed by an Amn lieutenant, that refers to instructions issued by "the respected Amn director," most likely the head of the General Security Directorate in Baghdad, on September 25, 1990. (MEW 102/7-A).

30 Contained in a letter of the Military Intelligence Directorate - Eastern Sector, ref. #168 on February 16, 1990. (MEW 2003/2-A). The order goes on to state ominously: "If such a case is found, the punishment will be severe." The "fate of the detainees" refers to standard practice in Iraq not to notify families of the arrest of their relatives, at least not until after court sentencing.

31 Memorandum of the General Security Directorate in Baghdad of February 10, 1988 (ref. Sh.3 Q.2 / 8970), circulated by the director of the Amn of Erbil Governorate on February 20, 1988, ref. #2629. (MEW 83/2-B). The desire to conceal the names of senior authorities is clear from this order, but initially, two interpretations are possible for the officials' motivation. One is that they did not wish to be held accountable at some future date for deeds that would constitute clear breaches of international law, like the mass killing that was about to take place. Another (suggested by the ambiguous phrase: "take their regular position") is that they did not want to be harangued by junior officers about the precise detail of directives. At the time of Anfal, official instructions were proliferating at a dizzying speed, and some were so strong (e.g., the order to execute all persons detained in the prohibited areas between the ages of 15 and 70) that they sowed confusion and uncertainty in the lower echelons of the security apparatus. We have found pieces of correspondence, for example, that ask for confirmation whether certain particularly severe orders should indeed be implemented.

The argument against this second interpretation is that junior officials in any case would not be permitted to address senior authorities directly, but would have to go through their immediate superiors in the regional offices. The issue addressed by the above directive is thus one of accountability for crimes.

32 Literally: "Please, (under)take what is necessary."

33 Letter from the Military Intelligence section of the Command of the Oil Protection Forces in Kirkuk, ref. #337 of April 15, 1988. (MEW 2110/9-A). The arrests took place at the height of the third Anfal operation in the Germian plain. (See Genocide in Iraq, chapter 5). The COPF's leader was Major-General Bareq Abdullah al-Haj Hunta, who commanded Iraqi forces in the third Anfal operation. The Security Directorate in al-Ta'mim Governorate, i.e., Kirkuk, is the last known place of detention of families arrested during Anfal. After this they "disappeared." Middle East Watch testimonies indicate that most were taken to sites in the western desert and killed.

The document sports a slogan in the top right-hand corner: "We Seek Justice, Not War." At the bottom, the numbers of detainees are given: sixty-nine women, one hundred children and ten men. (See also Document 23 in the Appendix below).

34 Literally: "security-prohibited areas," translated here usually as "areas prohibited for security reasons."

35 "Agents of Iran" is a reference to the alliance between the PUK and Iran, first consummated in October 1986. "Offspring of treason" refers to Masoud Barzani, whose late father, Mullah Mustafa Barzani, had been branded a traitor for fighting the central government since 1961.

36 Draft dodgers (mutekhallefin) who "surrendered" were sent to the local army recruitment center.

37 See Document 23 in the Appendix below.

38 Situation of human rights in Iraq. United Nations General Assembly, A/47/367/Add.1, 13 November 1992, p. 16.

39 Reproduced in U.N. Economic and Social Council, Report on the situation of human rights in Iraq, pp.

52-53.

40 For the authenticity of individual documents, see below at pp. 30-31.

41 Not all the documents are written on stationery. The files include many handwritten letters and even notes scribbled on regular sheets of paper. This is particularly true for Amn and Ba`ath Party offices in smaller towns, but not so much for the Istikhbarat. The reason is not immediately clear. Perhaps it was an issue of cost, since the stationery had to be brought in from one of the larger centers; or perhaps there was seen to be no need to write inter-office correspondence concerning matters of a purely local nature on official stationery.

42 Ibid., pp. 55-57.

43 Mission Permanente de la Republique d'Irak aupres de l'office des Nations Unies a Geneve, Response of the Iraqi Delegation to the Report presented by the Special Rapporteur of the Commission on Human Rights in Iraq at the 48th Session. (Geneva, February 20, 1992), p. 7.

44 Moreover, the documents are in fact written in excellent Arabic, clearly the product of native Iraqis who are fluent in the Iraqi dialect and its attendant vernacular. Not only that, the consistency of these documents also indicates that without a shadow of a doubt they were written by bureaucrats, and indeed by bureaucrats who were well-versed in the official discourse of the Iraqi state apparatus.

45 The PUK and KDP shipped most, but by no means all, of the documents in their possession to the United States for safekeeping.