1. Archaeological Report on Koreme, Birjinni, and Jeznikam - Beharke Cemetery
2. Forensic Anthropology Report
3. Report on Firearms Identification at the Koreme Execution Site
4. Convention on the Prevention and Punishment of the Crime of Genocide
5. Memorandum: The Elements of Crimes Against Humanity Applied to The Destruction of Koreme
_______________________
Acknowledgments
Maps
Plans and Diagrams
Photographs
Archaeological Report on
Koreme, Birjinni, and Jeznikam-Beharke Cemetery
by
James Briscoe,
Forensic Team Archaeologist1
and
Clyde Collins Snow,
Forensic Team Scientific Leader2
Archaeological investigations were carried out at three sites in Kurdistan, Northern Iraq, between May 25 and June 24, 1992 as part of investigations by Middle East Watch and Physicians for Human Rights ("MEW/PHR") into allegations of gross violations of human rights possibly amounting to genocide by the government of Iraq against Iraqi Kurds during the Anfal campaign of 1988. Archaeological expertise was considered an important part of these investigations in order to provide a precise record of events and remains at the sites investigated. These investigations were part of a multidisciplinary approach including forensic anthropology, oral history, ballistic analysis, and review of written documentation.
Archaeological investigations between May 25 and June 24, 1992 were limited to three sites, with the major emphasis on the village of Koreme, district of Mengish, Dohuk governorate, Iraqi Kurdistan, and a mass grave site at Koreme initially explored by Dr. Snow, forensicanthropologist and forensic team scientific leader, in February 1992. The second site was the village of Birjinni, district of Zawita, Dohuk governorate, Iraqi Kurdistan, selected because it was alleged to have been attacked with chemical weapons by the Iraqi army in August 1988. The third site was a cemetery at the Jeznikam-Beharke camp, near the city of Erbil, Erbil governorate, Iraqi Kurdistan, said to contain the remains of Kurds who died in the course of forcible relocations during the Anfal campaign of 1988.
Archaeological investigations included the general survey of the sites, mapping of salient features, controlled collection of artifacts and soil samples, and excavation of grave sites for forensic study.
This discussion incorporates by reference materials contained in the foregoing report, "The Destruction of Koreme During the Anfal Campaign."
The village of Koreme is located in a small valley of the frontal range of the Zagros mountains about four kilometers north of the district capital of Mengish in Dohuk governorate.
The Village Site
Koreme, prior to its destruction, consisted of two clusters of buildings divided by a small stream running north-south through the center of the village. The prominent landmark of the village is a 10 meter hill used as the original village cemetery. There were originally about 100 houses in the village, including 50 belonging to members of the Barwari tribe to the west of the stream and 40 to 50 belonging to members of the Shearali tribe to the east of the stream.
Houses in the village were generally about 5 x 8 meters in plan, with a few larger structures interspersed throughout the village. The average house was made of mud bricks, while the larger structures were made primarily from concrete blocks with limestone stem walls averaging about 30 cm.
The village school was located on the southeast edge of the village and was used as a reference point for mapping the village. (See Koreme Village Plan.) The school was a 15 x 20 meter two room building made of limestone and reinforced concrete. The village mosque was located onthe northwestern edge of the village and appeared similar to the school in size and design.
Koreme was bordered on the north, east, and west by agricultural lands, which in May and June 1992 were mostly fallow. Smaller garden plots were reopened in 1991-92 throughout the village, and one larger field was in cultivation about 800 meters west of the village. Forensic team investigators were shown a U.S.-made star shell on the south side of the field, rigged with a trip wire to one end; it is not known whether or not it was an operational booby trap, although local informants commented on problems with landmines in the fields. There is a large orchard on the south side of village and several smaller vineyards scattered about the area.
At the time of the forensic team's investigations in May and June 1992, all village structures, including houses, the school, and the mosque, had been destroyed down to the foundations. The mud houses appeared to have been bulldozed and, in some cases, possibly dynamited. The concrete and limestone structures appeared to have been dynamited.
The pattern of rubble scatter was significant in that it was primarily inside the structures and following neat lines, indicating interior charge placement with some care and expertise in demolition. Demolition appears to have been systematic, and there appears to have been no battle or other fighting resulting in artillery or bombardment damage to the structures. So far as could be determined, the destruction of Koreme was carried out by demolition and bulldozer, in a systematic fashion. The rubble of the village school contained a visible trip wire sticking out of the concrete remains, which local informants said was connected to a mine or booby trap.
The execution and grave sites are located across a small hill on the west side of the village, about 200 meters from the village proper.
The Execution Site
The execution site consists of the two lines along which, according to oral testimony and forensic evidence, the Iraqi firing squad stood and the Koreme victims squatted. (See Plan of Execution Site.) The execution site is on the steeper west slope of the hill and not directly visible from the village proper. A fallow plowed field in 1992 extends from the top of the hill to the upper edge of the execution site. According to local informants, this field was first plowed following the 1988 execution in 1991 and the edge of the execution site may have beenimpacted somewhat. The majority of the execution site, however, is located on a steeper slope that has not been plowed since the execution.
The firing squad, according to oral testimony and ballistic analysis, stood along a line uphill from the line where the Koreme men squatted. Slope grade is 10% between the two lines, which are about 12 meters apart. The Koreme victims' line is located about 8.7 meters southeast of the south edge of the grave site.
The execution site was divided into two sections for controlled collection purposes. A metric grid was established along the firing line and artifacts were collected in meter strips along the baseline. Each cartridge brass was plotted according to its position along the baseline and given a numerical designation before collecting. The numerical designation was plotted on graph paper and written on the artifact as each was located and collected. A second grid line, keyed to the firing line baseline, was set up along the victims line for collection of artifacts there. All materials located between the two lines and outside the grid were measured in relation to the grid baseline.
Artifacts were collected by two forensic team archaeologists searching meter wide strips on hands and knees. Koreme villagers assisted by cutting grass and weeds from each strip as it was searched and by marking the locations of artifacts as they observed them. Special care was taken not to move any artifacts until they were plotted and numbered.
Several artifacts were found besides spent brass. A shrapnel fragment was found on the firing line. A lighter, comb, tin cup, and a pair of man's shoes were found on the victims' line. Two unfired rounds were found between the two grids and two misfired rounds were found on the firing line.
Spent brass and all records of location and plotting were sent for ballistic analysis, results of which are found in Appendix 3.
The Grave Site
The primary focus of forensic team investigations was the mass grave site, dating, according to oral testimony, from the execution during the 1988 Anfal campaign. The grave site consisted of two low (approximately .75 meter high) concrete brick enclosures, each 5 x 8 meters in size, located about 2 meters apart on a slope west of the village.
The two low-walled areas were designated Graves A and B for control purposes and the two pits in each grave were designated GraveA-North, Grave A-South, Grave B-North, and Grave B-South so that each could be excavated separately. Bodies were given sequential numerical designations as they were exposed in each pit. (See Koreme Graves Plan.)
Each enclosure contained two roughly 2 x 2 meter pits. One pit, designated Grave B-S, contained several disturbances that appeared to be artillery shell craters predating the graves. Grave B-N included two similar disturbances and Grave A-S contained one. The four pits run along a line on the slope about 10 meters northwest of the line where the Koreme victims were killed. The site of the graves appears to have been based partly on the presence of existing holes, the shell craters. Shrapnel fragments were recovered from the shell crater disturbances. The grave site walls were made from 20 cm long concrete bricks.
Prior to actual excavation, the area within the two brick walls was cleared of brush and ground cover. The walls were then dismantled after mapping, to afford easier access to the graves and prevent collapse during excavation. the area around the grave site was also mapped, including the lines where the firing squad and Koreme victims had been.
The test made by Dr. Snow in February 1992 was reopened and expanded to cover the entire pit. Shovels were used to remove disturbed surface soils from the remainder of the grave pits before excavation commenced.
A "floating" grid system was established for each pit due to the habit of Koreme villagers of crowding around the excavations. Landmarks in each pit (pit outlines, skulls, other bones, etc.) were graphed for each of the drawings made and superimposed on a composite drawing in the laboratory. Photographs of each pit, landmarks, and body were periodically made for documentation and as a reference check for the composite map.
In the pits, excavation tools were confined to trowels, brushes, and bamboo picks. Standard professional procedures for excavating burials were followed at all times. Once remains were exhumation, the forensic team removed them following the system established by Dr. Snow in exhumations in Argentina. Artifacts, clothing, and skeletal remains were recorded on standard field inventory forms and removed from the grave pit in anatomical order (generally from foot to head in order). Evidence of trauma was noted on a skeletal checklist form as each bone was removed. All items were catalogued and bagged according to the numerical and pit designations assigned by the archaeologist in thefield. This information was transferred to Case File numbers assigned to each body as it was received in the morgue in the nearby city of Dohuk.
Grave A. Grave A was a walled enclosure measuring 4.5 x 8 meters, roughly the same size and shape as most of the houses in Koreme village. The grave contained two burial pits designated Grave A-N and Grave A-S. The grave was erected on a gentle slope, running northeast downhill to southwest and is oriented on its long axis 40 degrees west of magnetic north. Slope in the walled grave is 8 per cent northeast to southwest.
Grave A-N is a shallow 1.7 x 2.7 meter pit in the north corner of the walled grave. Two bodies were located between 0.3 and 0.5 meters below the surface on one side of the pit. A conical disturbance, 60 cm across, was located in the northeast corner of the pit and is believed to have been an artillery shell crater, as suggested by shrapnel found in other similar disturbances.
Grave A-S is a larger and deeper pit, 2.8 x 3.2 meters, located in the southeast corner of the walled grave. The pit has an irregular outline and appears to have been expanded on one side prior to filling. Eight bodies were recovered from the eastern two-thirds of the pit below 0.4 to 0.8 meters of fill. The southwestern one-third of the pit contained 0.1 to 0.2 meters of fill and may have been too shallow for covering bodies.
Grave B. Grave B is located just southeast of Grave A and in a 5.2 x 9.1 meter enclosure oriented on its long axis 20 degrees west of magnetic north. The southwest corner of the wall is roughly 10 meters northwest of the line where the Koreme victims were squatting during the execution. Slope within the walled grave runs downhill east to west and is 5 per cent. There is a small almond tree in the center of the south half of the enclosure, planted by one of the local villagers in 1985, that served as the datum point for the grave site. Grave B also contained two burial pits, designated Grave B-N and Grave B-S.
Grave B-N is a 2.8 x 3 meter pit located in the northeast corner of the enclosure. Eight bodies were located in the pit between 0.1 and 0.35 meters deep. Two roughly 60 cm wide disturbances were noted on the east edge of the pit that contained a powdery white substance and shrapnel fragments and are believed to be artillery shell craters that existed prior to the grave pit. Bodies in the pit appeared to be more tightly packed and commingled than in the Grave A pits.
Grave B-S was located in the southeast corner of the enclosure and measured 1.9 x 2.35 meters. Nine bodies, tightly commingled, were found between 0.2 and 0.4 meters deep. Three roughly circular conicaldisturbances, 60 cm across, were found on the edges of the pit and are thought to be artillery shell craters predating the burial. At least four other similar disturbances were also noted within the Grave B enclosure but were not examined. One disarticulated humerus was found in the 0.2 meters of fill above the bodies.
The purpose of the visit to the village of Birjinni, district of Zawita, Dohuk governorate, on June 10, 1992, was to obtain data concerning chemical weapons attacks occurring in late August 1988.
The Birjinni Site
The village of Birjinni occupies a narrow saddle and mountain pass along a chain of higher ridges between the towns of Zakho and Dohuk. It comprises a tell 10 meters high and 100 meters wide. There is a low sloping terrace on the north side of the village, consisting of about 0.3 hectares with orchards and limited farmland. (See Plan of Birjinni Village.)
Birjinni had about 40 houses at the time of the attack in August 1988, some of stone and others of mud brick. It also had a school and mosque, each a two room building made from cement and stone. There were remains of two roads to the village, one leading to Dohuk and the other to Zawita. At the time of the June 1992 investigations, only the Zawita road was passable, and there were posted landmine warnings adjacent to the road. In June 1992, no buildings remained standing in Birjinni. The sites of the mosque, school, and houses had been destroyed down to the foundations, in the same fashion as that described in Koreme.
The Chemical Weapons Bomb Craters
Four bomb craters along the west edge of the terrace, about 700 meters from the village, were examined in detail. Visual confirmation of the locations of the other eight craters was also made.
The four craters examined in detail consisted of low conical depressions 2.2 meters across and 0.6 to 1.2 meters deep. Fragments of the bombs were found lying immediately beside and in the craters, andin two instances consisted of an iron outer envelope, an aluminum inner canister, a heavy lid labelled "Top" in English, a spout in the lid, and twisted tail fins. The fragments near each crater in those two instances were sizable, approximately 1 meter by 0.5 meter by 0.5 meter, and approximately 10 kilos in weight. Small bits of a yellowish ocher-like substance were noted in the craters and scraped from inside a canister and, along with soil samples, have been sent for laboratory testing. The four craters were evenly spaced on a straight line thirty meters apart and may have been dropped from low altitude, on a line consistent with survivor reports of aircraft direction.
The Exhumation Site
The forensic team carried out exhumations of two reported chemical weapons victims buried on adjacent plots on the sloping terrace lying north of the village, down valley and near a stream flowing down from the saddle. The site of the exhumations was a wooded canyon, filled with both cultivated fruit and wild trees. The skeletal remains were removed from graves each approximately one meter deep. Standard procedures for exhumation and examination were followed, and the results noted in Dr. Snow's anthropological report. The two skeletons were reburied in the village cemetery following forensic examination in accordance with Islamic ritual.
Jeznikam-Beharke cemetery, reported to contain the remains of forcibly relocated Kurds who had died after being brought to these camps outside Erbil, was examined between June 18 and 20, 1992. Archaeological investigations included a sample inventory of the camp graves compared to graves in the previously existing cemetery that had served the old village prior to establishment of the 1988 camps. Forensic examinations of three children's graves in the camp portion of the cemetery were carried out by Dr. Snow on site. The remains were replaced in their crypts once investigations were complete.
The Site
Jeznikam-Beharke cemetery is located on a small conical tell about ten meters high and 135 meters across. An older existing cemetery is located atop the tell and is associated with a nearby village that was destroyed by the Iraqi army in 1987 in the course of establishing the Jeznikam-Beharke camp.
The dead of Jeznikam-Beharke camp were reported by camp survivors to be buried on the southern and eastern slopes of the cemetery. The camp portion of the cemetery covers an area roughly 30 x 101 meters on the southern edge and 10 x 100 meters on the eastern edge.
Sample Inventory of Graves
Methodology. The forensic team's analysis accepted the proposition that, in populations under extreme stress, there would be not only an increase in overall mortality but also a shift toward a proportionally higher death rate among children. It also seemed reasonable to assume that the graves in the village part of the cemetery represented a fair picture of the long term mortality profile of Jeznikam village which, prior to Anfal, was a typical Kurdish community of its size. In such a community, we would expect the ratio of child to adult deaths, taken over many years, to be more or less constant. This ratio would be reflected in the proportion of child to adult burials in the village cemetery. It could therefore serve as a "control" with which to compare the corresponding ratio of the detainee section of the cemetery.
Kurdish burial custom requires that the dead be stripped of all clothing and adornments, washed, and wrapped in a plain linen shroud. Graves are normally dug to a depth of about 1.8 meters. The bottom, lower sides and ends of the grave are loosely lined with flat stones to form a crude coffin-like crypt. The fully-extended corpse is placed in the crypt with its head at the west end of the grave and lying on its left side so that it faces southward toward Mecca. After the body is properly positioned, the crypt is sealed with another layer of flat stones placed over the corpse and the grave is filled.
With few exceptions, the graves of Jeznikam offer no information on the identity of the dead. Following their tradition of extremely simple burial, Kurds do not ordinarily inscribe gravestones with the decedent's name, age, or date of death. However, because a grave is dug no longerthan necessary comfortably to accommodate the corpse, its length is roughly proportional to that of the body. Stated more simply, the graves of children are shorter than those of adults. It follows that, within a given series of burials, the ratio of shorter to longer graves should provide an approximation of child mortality relative to that of adults. Therefore, the forensic team decided to use grave length as an index by which to compare the child mortality profile of the Jeznikam villagers during normal times with that of the detainees during their confinement in Jeznikam-Beharke camp.
A complete census of graves in the two sections could not be undertaken in the brief time available. Instead, the forensic team conducted a sampling survey based on a series of seven equally-spaced transects. Each transect spanned the entire breadth of the cemetery from north to south and thus included areas of both the village and detainee sectors. They were also perpendicular to the axes of the graves which are oriented from east to west. If a transect line passed between the head and foot stone of a grave, it was included in the sample and its length and sector recorded.
"Grave length" was defined as the distance to the nearest .05 meters between the top centers of the head and foot stones. It is somewhat longer than the actual length of the corpse, on account of the fact that the crypts are constructed to allow some free space between end walls and the corpse, and further because of the thickness of the stones forming the ends of the crypt. Based on previous observations of Kurdish burials, this difference appears to average about 30 cm. Accordingly, this amount was subtracted from measured length to obtain a truer approximation of "body length," which was to be used as the principal variable in our data analysis.
Since the purpose was to compare the proportion of children's graves to those of adults, it was necessary to develop a reasonable criterion for distinguishing them based on estimates of body length. To do so, the forensic team utilized Field's anthropometric survey of Iraqi Kurds.3 From Field's data -- the most extensive to date -- it appears that a range between 141 cm (female mean - 2SD) and 178 cm (male mean + 2SD) would embrace about 96.5% of Kurdish adults of both sexes. Based on this, the forensic team classified burials shorter than 141 cm. as "subadult" and those 141 and longer as "adult." There would inevitablybe some overlap on account of exceptionally short adults or larger adolescents; these two categories would tend to offset each other.
Results. A total of 166 graves fell within the sampling parameters. Of these, 81 (48.8% were in the village sector and the remaining 85 (51.2%) were in the detainee sector. In the village sector, subadult burials comprise 45 of the total of 81 sampled burials or by a ratio of 1.25. Such a finding is not unexpected in a peasant society where childhood mortality from epidemic disease is high and health care not good. In contrast, subadults in the detainee sector comprise 71 (83.5%) of the sample 85 burials. In the detainee sector, the subadult to adult ratio is 5.07. A Chi-square analysis of the distribution shows that the difference between the two sectors is statistically significant at the .0001 probability level.
However, the very high subadult to adult ratio of graves in the detainee sector cannot be attributed entirely to a higher mortality rate among children. A second factor influencing the ratio relates to the Iraqi government's elimination of substantial numbers of adult males either by execution or forcible disappearance prior to the transfer of remaining detainees to the camp. For this reason, the adult component of the detainee population had already been reduced by the time the detainees had reached Jeznikam. This fact has the effect of making the subadult to adult grave ratio in the detainee sector less statistically unusual. To assess this effect, the forensic team calculated the distribution of subadult and adult graves that would be expected if two-thirds of the males had been eliminated from the detainee population prior to the detainees' arrival at Jeznikam. When this hypothetical distribution is compared to that actually observed in the detainee sector the difference is still significant at the .05 probability level.
Thus, it is evident that a disproportionate number of deaths occurred among children of the detainees, as substantiated by the ratio of subadult to adult graves in the detainee sector of the cemetery. This ratio, which amounts to about five subadult graves to one adult grave, is significantly higher than that observed in the village sector, which was taken as representative of the normal mortality profile among rural Kurds.
* * *
The investigations referenced in the foregoing were carried out according to accepted professional archaeological and anthropological standards, and the statements in the above report are true and correct to the best of our knowledge and belief.
Respectfully submitted,
James Briscoe
Forensic Team Archaeologist
Clyde Collins Snow
Forensic Team Scientific Leader
Summary of Anthropological Report
by
Clyde Collins Snow,
Forensic Team Scientific Leader
The following discussion summarizes exhumations undertaken by the forensic team in Koreme, Birjinni, and Jeznikam-Beharke cemetery. The full scientific and forensic report on each exhumed skeleton may be obtained from MEW/PHR and will be submitted as evidence to any judicial tribunal hearing charges based on the foregoing report, "The Destruction of Koreme During the Anfal Campaign," the contents of which are incorporated by reference into this discussion.
The forensic team exhumed 27 skeletons from two graves at Koreme. All 27 skeletons were male, ranging in age from early adolescent to approximately early 40s. All appeared to have suffered death by gunshot wounds.
In most of the 27 cases, detailed study of the fracture patterns in anatomical relation enabled the forensic team to determine the number of wounds suffered by each individual. Wherever possible, the forensic team also tried to establish the trajectory of the projectile.
As a first step in this analysis, fragmented bones were reconstructed with the aid of a hot glue gun. Reassembly of the thorax and vertebral column of each skeleton, also by gluing, was particularly useful in establishing trajectories of wounds involving these regions. Bullets or their fragments still embedded in the bones also provided certain clues. Bullet holes in the victims' garments, studied in relation to the observed osteological trauma, gave clearcut evidence of trajectories in many cases. These findings on a per skeleton basis are available in the full anthropological report.
In some cases, it was impossible -- due to the complexity of the fracture patterns -- to determine whether the trauma was to due to a single projectile or several. Thoracic wounds with widespread and multiple rib fractures are particularly apt to be ambiguous in this respect because it is so often difficult to determine whether the trauma was caused by more than one bullet or by fragments of a single bullet. In our analysis, such cases were classified as a single wound. Because of this, our final tabulation of the number of wounds observed may be an underestimate. Another source of possible error is due to the difficulty of diagnosing compound wounds. For example, from skeletal evidence alone, it may be impossible to determine whether a bullet passed through the forearm and continued to enter the chest. Therefore, some of the wounds we have included in our tabulation may represent compound wounds caused by a single bullet passing through two or more body segments. Finally, it must be recalled that a bullet may cause wounds without any discernible osteological trauma at all. It is not uncommon, for example, for projectiles to cause through-and-through abdominal wounds without striking bone.
In all, there was clearcut evidence of at least 84 separate wounds distributed among the 27 skeletons. Five, or 18.5%, of the series showed evidence of a single wound. In the remaining 22, the number of wounds ranged from two (seven individuals) to six (one individual) with an average of 3.1 wounds per individual.
The distribution of the identified wounds by body region shows that wounds of the upper trunk (thorax and shoulder) were most prevalent, making up 28.6% of the 84 observed. Fifteen wounds involved the pelvis (including the lower lumbar vertebra). Thus, in all, wounds of the trunk (thorax and pelvis) numbered 39 or nearly half (46.4%) of those observed. Wounds to the extremities made up 41.7% of those observed; there was no statistically significant difference in their distribution by region (15 were in the upper extremity and 20 in the lower) or side (15 left, 20 right). As pointed out above, many of these extremity wounds --especially those of the upper arm and thigh -- were probably compound. Ten (11.9%) of the wounds were to the head.
Each of the identified wounds was studied to determine its trajectory. A given trajectory can be defined in terms of its directional components in relation to the principal anatomical axes. In two cases, trajectories were indeterminate, reducing the number of observations to 82 from the total of 84 wounds.
The distribution of the 82 observed wounds by direction shows the results of Chi-square tests on the data. Wounds from projectiles entering the front of the body and those entering from the back were nearly equal in number. The same is true for those entering from the right and the left. However, wounds with downward trajectories occurred over twice as frequently as those from projectiles passing upward and this difference is strongly significant (p < .001).
The anatomical distribution of wounds observed in the 27 victims suggests a random and indiscriminate firing pattern with the principal aiming points in the trunk. This is also supported by the anatomical distribution of the observed wounds. Of particular interest is the low frequency of head wounds, suggesting that the traditional coup de grace, in the form of a single gunshot wound to the head, was not systematically rendered in this case.
Trajectories are also revealing. For example, the statistically significant preponderance of downward-passing bullet paths indicates that the shooters fired from a position somewhat higher than the victims. although bullets entering from the front of the body and from the back caused about an equal number of wounds, in almost all cases the transverse was the predominant vectorial component. Stated more simply, most of the wounds entered the side of the body rather than directly from the front or the back. This finding indicates that the victims were unrestrained and, perhaps reacting to visual stimuli (raising of firearms to a firing position) and/or auditory (a command to commence firing) cues, had already begun involuntarily to turn or twist away from the executioners when first struck.
In summary, the ballistic and wound evidence is consistent with the accounts given by survivors and other witnesses.
Investigations were undertaken in Birjinni for the limited purpose of determining whether forensic data was consistent with villagers' accounts of chemical weapons attacks in August 1992 resulting in four deaths and an undetermined number of injuries. The purpose of exhumations was to find out whether any traces of chemical agents remained on the clothing of victims after four years of burial, in the case of victims who had been buried unwashed and in their original clothing relatively soon after death from chemical agents. Birjinni was thusselected as the investigation site on account of survivor reports that two victims, an old man and a boy, had fallen victim to chemical agents and buried shortly thereafter in their original clothing.
The forensic team exhumed skeletons from two adjacent graves at Koreme, following standard procedures. The first skeleton was determined to be that of an old man, of approximately sixty years. Surviving family members identified the remains as those of the grandfather, on the basis of artifacts and clothing. The second skeleton was that of a young boy, of approximately five years. Family members identified the remains as those of the son, on the basis of clothing. Examination of the skeleton revealed no signs of trauma or perimortem violence, or any other indication inconsistent with accounts of the chemical weapons attack given by surviving villagers. Laboratory analysis has thus far been unable to detect any signs of chemical agents in clothing, soil, or bone samples.
The forensic team undertook three exhumations in the detainee sector of Jeznikam cemetery for the purpose of determining whether forensic evidence would reveal any indications inconsistent with accounts of detention, and resulting privation and death, described by surviving detainees. The forensic team was accompanied in its Jeznikam investigations by a Koreme villager who said he had buried his mother and infant sister, Farwan Tawa Mostafa, in the detainee sector of the cemetery.
The forensic team exhumed at the site where the Koreme villager indicated he had buried Farwan. The team encountered the skeleton of a female infant, interred in a dress identified by the Koreme villager as the one in which he had buried his sister.
The mandibular central incisors of the skeleton were almost completely erupted and the maxillary centrals unerupted, suggesting an age at death of around seven (plus or minus four) months. However, using diaphysial length of the long bones as an aging criteria, Farwan would have been classified as no more than 1-3 months old. This marked discrepancy between dental and skeletal age is diagnostic of severe nutritional and/or disease stress since, in such cases, dental maturation tends to remain in step with chronological age while skeletal growth is severely retarded. The bones displayed no signs of perimortem violence.
Accordingly, the forensic team finds that Farwan Tawa Mostafa was a female infant likely suffering from severe malnutrition or disease. The forensic team found no evidence inconsistent with accounts given by detainee survivors.
* * *
The investigations summarized above were carried out in accordance with accepted anthropological standards. A complete report, including data for each exhumed skeleton, is available from MEW and PHR. The statements and conclusions above are true and correct to the best of my knowledge and belief.
Respectfully submitted,
Clyde Collins Snow,
Forensic Team Scientific Leader
Firearms Identification
of the Koreme Execution Site
by
Douglas D. Scott, Ph.D.
Lincoln, Nebraska
The cartridge cases, totaling 124, from the Koreme execution site were analyzed to determine the minimum number of shooters. Sixty-three cases were collected from the site surface and their location was piece-plotted. The piece-plotting proved valuable in ascertaining movement of individual guns during the execution. These cases were identified by a number (1 to 63) written on the case body in indelible ink.
Seventeen additional cases were recovered during the excavation of Grave B-S. They were separately bagged when received. A letter (A through Q) was arbitrarily assigned by the author. The letter was written on the case body with a fine point Sharpie. A final group of cases totaled 44. These were recovered in pile near an olive oak tree about 20 meters north of the piece-plotted cartridge cases. For convenience these cases were arbitrarily numbered (100 through 143) with a Sharpie.
Methods of Analysis
The comparative study of ammunition components is known as firearms identification analysis. Firearms, in their discharge, leave behind distinctive metallic fingerprints or signatures on the ammunition components. These signatures, called class characteristics, allow the determination of the type of firearm (i.e. model or brand) in which a given cartridge case or bullet was fired. This then allows determination of the number of different types of guns used in a given situation.
Further, they allow the identification of individual weapons by comparing the unique qualities of firearm signatures, individual characteristics. This capability is very important because coupled with the precise artifact locations, identical signatures can be used to identifyspecific firing areas. With this information, patterns of movement can be established and sequences of activity can be more precisely interpreted.
The means to this end is reasonably simple in concept. When a cartridge weapon is fired the firing pin strikes the primer contained in the cartridge, leaving a distinctive imprint on the case. The primer ignites the powder, thus forcing the bullet down the barrel. The rifling in the barrel imprints the lands and grooves on the bullet in mirror image. The extractor also imprints the spent case as it is removed, extracted, from the gun's chamber. These imprints are called individual characteristics.
Police agencies have long used the investigative technique of firearms identification as an aid in solving crimes. Two methods commonly used by police departments include comparisons of bullets and cartridge cases (Harris 1980; Hatcher, Jury, and Weller 1977) to identify weapon types from which they were fired. Police are routinely successful in matching bullets and/or cartridge case individual characteristics to the crime weapon simply by demonstrating that the firing pin, extractor marks, or the land and groove marks could only have been made by a certain weapon. In the event that weapons used in a crime are not recovered, police can say with certainty, on the basis of the individual characteristics, from recovered bullets and cases, that specific types and numbers of weapons were used.
The comparison microscope is critical to the analysis of ammunition. Simply, the microscope is constructed so that two separate microscope tubes are joined by a bridge with prisms mounted over the tubes. Two separate images are transmitted to the center of the bridge, where another set of prisms transmit the images to central eyepieces. The eyepieces are divided so that each image appears on one-half of the eyepieces. Movable stages allow the objects under scrutiny to be manipulated so that they can be directly compared for class and individual characteristics.
The microscope used in this analysis was a Bausch and Lomb comparison microscope. The objectives range from 10 to 50 power. Each cartridge case was examined to determine its class characteristics. The cases were all fired in a 7.62-caliber firearm. The cases appear to have been fired in a semi-automatic or full automatic gun like the AK-47. All cases were fired in similar type firearms.
Following class characteristics determination the cases were compared to one another to determine individual characteristics. When possible matches were identified these cases were set aside until thesequence of initial identification was completed. Then each group was re-analyzed. Once a case group, arbitrarily numbered one through seven, was verified, two to three cases from each group were then compared to the other groups to further cross-check identification validity.
Results of Analysis
The firearms identification analysis indicate there were at least seven individual firearms used in the execution. The firearms were all semi-automatic or full automatic 7.62 x 39 mm caliber firearms. Multiple matches were made with all cases. The seventeen cases found in the grave and the 44 found near the olive tree matched to the piece-plotted cases. The firearms evidence strongly suggests only a single event involving the firing of over 100 rounds of 7.62mm caliber ammunition occurred at this site. Those involved in the shooting are minimally identified as seven individuals.
Table 1 identifies the case groupings that match. The matching groups were arbitrarily numbered one through seven. Among the cases are five unfired rounds (1, 15, 17, 62, and 63) found intermixed with the fired rounds. A single cartridge (number 16) was a misfire. Individual number 4 fired at least 15 rounds. Some time during the shooting he had a round misfire that required it be cleared from the chamber by manually working the firing bolt. The condition of his gun's firing pin, as seen in the imprint on the primer, strongly suggests this gun was very dirty and possibly in poor condition.
The AK-47 and similar model firearms have a detachable magazine that contains 30 rounds. Assuming each shooter had loaded a 30 round magazine in his weapon prior to the execution it appears that individuals 1, 2, 3, 4, 5, and 7 fired at least one partial magazine each. Recovered cases indicated each minimally fired between 12 and 17 rounds, which is approximately one-half of a full magazine. Individual number 6 fired at least 37 rounds. This individual had to reload at least once during the execution.
The distribution of the cases, as piece-plotted shows two distinct clusters of cases. One group of six or seven cases lies in a western cluster. The second and largest group is 16 meters to the east. A possible third cluster lies in a rough line to the north of the western group. There is a gap of about 16 meters between the eastern and western clusters where only a single case, number 11 was found.
When the case matches are plotted it becomes clear that the separation of cases is more apparent than real. Five individuals (numbers 3, 4, 5, 6, and 7) fired in the eastern and western clusters. The plotted matches in the western cluster suggest the individuals were aligned in a somewhat linear arrangement in this area. The eastern group demonstrates a much more bunched grouping.
I speculate that the firing squad may have been a linear arrangement on the west and as the firing began the men moved to the east in a random manner. The absence of cases between the west and east may be artificial in that the cases found in the grave and piled near the tree may imply someone picked the central area nearly clean.
The case distribution and matches make it clear that individuals number 2, 4, 6, and 7 moved to the north from the western cluster and fired one or more rounds each as they neared the victim line. Individual number 6 fired at least 12 rounds as he moved toward the victim line. At least two rounds were fired within ten meters or less of the victim line by individual number 6.
In conclusion the cartridge case firearms identification analysis from the Koreme execution site indicate at least seven individuals were responsible for the shooting. The collection and piling of a large quantity of cases has undoubtedly disrupted the overall pattern so all conclusions presented are subject to this bias. Six of the seven individuals fired at least one partial 30 round magazine during the execution. One individual fired one whole and at least one partial magazine during the shooting. Individual number 6, who fired the most rounds, also moved the closest to the victim line as determined from the piece-plotted cases.
Group Matching Cases by Number
1. 2, 4, 6, 8, 9, 20, 23, 49, 106, 109, 111, 137
(total 12)
2. 3, 7, 24, 28, 31, 36, 41, 54, 113, 128, 140, 141
(total 12)
3. 5, 25, 29, 32, 34, 37, 40, 47, 48, 112, 120, G, J,
(total 13)
4. 10, 12, 14, 16 (misfire), 59, 102, 110, 121, 126, 127,
D, F, H, L, N, (total 15)
5. 13, 18, 19, 21, 22, 104, 114, 138, A, B, C, I, K, M, O,
P, Q (total 17)
6. 11, 26, 44, 45, 46, 50, 51, 52, 53, 56, 57, 58, 60, 61,
100, 101, 103, 105, 107, 108, 115, 116, 117, 118, 119,
122, 123, 124, 125, 129, 130, 131, 132, 134, 136, 139,
142 (total 37)
7. 27, 30, 33, 35, 38, 39, 42, 43, 55, 133, 135, 143, E
(total 13)
Harris, C.E.
1980 Sherlock Holmes Would Be Impressed. American Rifleman
128(5):36-39, 82.
Hatcher, Julian, Frank J. Jury, and Jac Weller
1977 Firearms Investigation, Identification and Evidence.
Harrisburg, Pa.: Stackpole Books
1 Mr. Briscoe is a field archaeologist with Roberts/ Schornik & Associates, Inc., Norman, Oklahoma, to whom MEW/PHR express thanks for making Mr. Briscoe available for the extended length of the dig in Iraqi Kurdistan.
2 Professor of Anthropology, University of Oklahoma, Norman, Oklahoma.
3 Field, The Anthropology of Iraq, Harvard (1952).