(Courtesy Department of Special Collections, McFarlin Library, University of Tulsa).


Confirmed Deaths:

A Preliminary Report

By Clyde Collins Snow


A Cautionary Foreword

It should be emphasized that this report is, as indicated in the title, preliminary. While collecting data for this study, it has become obvious that much critical information on how many people were killed and who they were is lacking. Much of this information still resides in the memories and family records and other personal documents of the survivors and participants of the riot - both black and white -- and their descendants. For this reason, we are reaching out, both locally and nationally, for more information on possible persons killed in the riot whose deaths were never recorded. We also suspect much additional information of importance is contained in still unexamined documents such as life insurance claims, will probates, census records, etc. Hopefully, these documents still survive in obscure archives. Until this data is collected and analyzed, no final report can be completed.


In my final report, I will include a full list of the many persons who have helped me. In this preliminary effort acknowledgments must be limited to the wise and indefatigable Mr. Dick Warner and Ms. Sue Bordeaux of the Oklahoma State Department of Health. Much of the basic information upon which this report is based was originally compiled by Dick; he is also a magnificent fact-checker. Sue Bordeaux's vast knowledge of the vital records system and her enthusiasm in putting it to work in this project was invaluable. Naturally, neither one of them are responsible for any factual errors or eccentric opinions which may appear in this preliminary report -- they are all my own.


From the first shots that were fired at the courthouse on May 31, to the last fighting that took place on June 1, the Tulsa race riot proved to be a particularly lethal affair. And while a definitive death count is still elusive, it is clear that dozens of blacks and whites lost their lives in the catastrophe (Department of Special Collections McFarlin Library, University of Tulsa).


The Need for Accurate Casualty Counts

During the past half-century, it has become increasingly common for major disasters, natural and man-made, to become the subject of public investigation. Such investigations may be official -- that is, conducted by any governmental branch, judicial, executive or legislative and at any level, federal, state or local. Unofficial, but no less searching and revealing, investigations may be conducted by the press or private entities. Examples of such inquiries in the recent past include the several investigations of the deaths of the followers of David Koresh in the Branch Davidian Compound in Waco, Texas in 1994. Such inquiries are designed to shed light on the causes of such disasters, establish culpability when possible and appropriate and provide guidelines to prevent or, if they do occur again, design procedures for effectively dealing with them in the future. When conducted objectively, they generally attain these goals.

Unfortunately, no impartial investigation was conducted of the 1921 Tulsa race riot in its immediate aftermath, while memories of the participants and victims were still fresh, and the physical evidence, including the bodies of the dead, could be forensically examined. Today, eight decades after the event, only the documentary evidence -- much of it lost or of doubtful authenticity -- and the fading memories of the rapidly dwindling survivors remains.

A key piece of information in any investigation of incidents involving loss of human life is an accurate assessment of the number of victims. Such determinations are important for several reasons. For example, where preliminary estimates of the number of dead are part of an ongoing investigation, they can be used to make reasonable allotments of often scarce manpower, equipment, and financial resources to the task and to determine the overall investigative strategy.

Accurate estimates of the dead and injured can also help identify factors contributing to such disasters and, thus, provide guidelines for ameliorating the loss of life in similar future cases. For example, in Honduras, the immigration of the rural poor to urban areas resulted in large numbers of them building small houses on "waste" land along the steep banks of major river courses and other areas subject to flooding. As a consequence, many thousands of such settlers drowned or died in mud slides during the massive hurricane of 1998. This loss of life could be minimized by governmental or private aid to provide housing sites in safer areas or, at the least, assure the prompt evacuation of people from such vulnerable places when warnings of impending hurricanes are received in the future.

When the disasters are man-made, such as acts of terrorism, war crimes or other massive human rights violations, an accurate assessment of the number of victims is a necessary step in any forensic investigation conducted to exhume the victims so that they may be identified and returned to the families, make suitable reparations to the persons affected and, hopefully and above all, provide evidence to bring the perpetrators to justice.

Before the ashes of Greenwood had cooled, disagreements over the number of dead began to surface. Estimates of the total number of dead have varied by an order of magnitude, ranging from about fifty to as many as five hundred. They also vary greatly in the reliability of the sources on which they are based. Here, I have chosen a more conservative approach by compiling a list of persons who have, at one time or another, been named as victims of the Tulsa race riot. At the outset, I should point out that this compilation is not likely to include all of the riot fatalities since it is probable that at least some and, perhaps many, deaths went unrecorded. At the same time, however, I feel that it may prove valuable to future scholars since it provides at least a firm minimum of the number of dead.

Classification of Deaths

Based on the information presently available, riot fatalities of both races can be divided into two groups. Within the first are those established by primary sources such as death certificates and mortuary records. The second group consists of deaths mentioned only in secondary sources (newspaper stories, magazine articles, books, etc.) dealing with the race riot. In this study, I have designated individuals in the first group as confirmed, and those of the second as reported deaths.

The distinction between the two groups is made clearer when put in a forensic context. For example, bearing in mind that there is no statute of limitations on murder and that the victims killed in the Tulsa race riot were homicide victims, it is at least theoretically possible that murder charges could be brought against an alleged perpetrator.1 If the victim were to be Dr. Andrew C. Jackson, the prominent black physician who was gunned down after emerging from his burning Greenwood home with his hands held high, the death certificate signed eighty years ago would be unchallengeable evidence of his death in any court.

On the other hand, let us imagine that an elderly black man was charged with the death of a white woman identified only as "Mrs. Deary" by the now extremely aged ex-Sergeant Esley of the Tulsa National Guard. Assuming that his story had not changed since it was recounted in the Muskogee (OK) Phoenix in 1921, Sergeant Esley would testify that the victim died in her husband's arms after being struck by five bullets fired by a black who stole up behind her while she and her family were watching the fires in Greenwood from the front porch of their home on Sunset Hill. He might further state, as he did eight decades ago, that, after watching his mother die, Mrs. Deary's fifteen year old son joined the riot and helped set some of the fires. On cross- examination, of course, Sergeant Esley would be forced to admit that even in 1921, when he first told his story, be bad not been able to remember the victim's name but only . . . "that it sounded like Deary." Furthermore, he was not sure whether she was shot late Tuesday night or on Wednesday morning. Now suppose, that the astute defense lawyer introduces (as they always do, at least on television), a "surprise" witness, and a fragile little old lady makes her way to the stand.2 She would state that her name was Mrs. S. A. Gilmore and that, in 1921, she was living at 225 E. King in the Sunset Hill addition, which overlooked the Greenwood district. On Wednesday morning, while she and her husband were watching the battle below, she received five wounds in the arms and chest. While the shots came in the direction of Greenwood, it was never certain whether they were fired by a black or she was struck by stray shots being fired in the general direction of Sunset Hill by members of the white mob. Taken to Morningside Hospital, she lingered close to death for several days3 but eventually recovered. The defense attorney would then introduce as documentary evidence Tulsa City Directories which show that Mrs. Gilmore did indeed reside at 225 E. King at the time of the riot in 1921 and, in fact, was stilling living there two years later. He would also point out that Mrs. Gilmore was the only white female reported to have been shot during the riot in the abundant local and national press coverage. And finally, he would show that an exhaustive search of death records failed to produce any evidence of the death of Mrs. Deary in the form of funeral home, cemetery or, most importantly, a death certificate. While the jury would rush out to acquit, the red-faced prosecutor would sit contemplating how much he would enjoy ripping out the pacemaker of his star witness, Sergeant. Esley.

The hypothetical trials for the murders of Dr. Jackson and Mrs. Deary, by juxtaposing the tragic and the comic, serve to illustrate the crucial difference between confirmed and reported deaths as I have classified them here. Only the most dim-witted prosecutor would consider actually taking the Deary case to court based on Sergeant Esley's story. On the other hand, the Jackson murder would have been a strong case for the prosecution since the documentary evidence clearly establishes his death and the witnesses, both black and white, could have provided clear and convincing evidence of the circumstances of his death. Unfortunately, however, no investigation of this death was ever undertaken by the Tulsa police or other city, county, or state officials.

Readers should be aware the categorization of individual deaths as confirmed or reported in this preliminary study is not necessarily final. This is because the data presently available on many of the victims is still incomplete. As further information comes to light, at least some of the deaths classified as reported might be fully confirmed. This is well-illustrated by the case of Ed Lockard, which will be discussed in detail in the final report.

As noted above, much more data must be collected and analyzed to produce a final report. This is particularly true in regard to reported deaths. Therefore, in this preliminary report, only the data so far compiled on confirmed deaths will be presented.


Analytic Method

The initial effort of this study consisted of combing all known documentary sources for the names of individuals mentioned as victims or possible victims of the riot. The most important primary source was, of course, contemporary local and national press accounts in which the names of riot victims were given. These names include not only the reported fatalities but, also, those who were wounded severely enough to be admitted to local hospitals. In addition to press stories, the various books, reports, and articles published in the years since the riot also were a source of names.

The next step in this analysis was to enter the names, along with other data pertaining to the victims, into a computerized database. Once entered, other information on a particular victim could be pursued. For example, an especially important procedure was to search for the person's death certificate in the files maintained by the Oklahoma State Department of Health, census data, Tulsa City Directories. Funeral home and cemetery records of the period also were helpful, and in a few cases, valuable information was supplied by the victimís family members.

Death Certificates

In 1921, Oklahoma death certificates consisted of two sections, one to be completed by the undertaker and the other by the physician who attended the deceased. Normally, the completion of a death certificate required four steps:

1. The undertaker would begin the process by filling in the personal data on the dead person. This would include the name, sex, race, age, occupation, birthplace and occupation of the deceased as well as the names and birthplaces of his or her parents. The informant (usually the next-of-kin) providing this information also was asked to sign the certificate.

2. The certificate would then be sent to the attending physician who provided the date, time and cause of death. Signed by the physician, it was returned to the undertaker.

3. Next, the undertaker would complete his part of the certificate by listing the cemetery and date of interment or, if the body was buried elsewhere, the date and place of shipment.

4. Finally, the undertaker would submit the completed certificate to the vital statistics registrar of the county in which the death occurred. After assigning it a unique register number, the registrar would forward it to the Bureau of Vital Statistics of the Oklahoma State Health Department in Oklahoma City.

In the case of the riot victims, the orderly process outlined above was not always followed. In particular, the personal information on the deceased was sometimes left vague or incomplete. Informants who were not immediate family members did not often know such details as the exact age, marital status, or birthplace of the deceased, much less the names of the dead person's father or mother. This was especially true for black victims since their next-of-kin were still in the detention camps and could not come to the mortuaries to claim their relatives if, indeed, they were informed of their deaths at all.


Death certificate for an unknown African American


The information provided by physicians also was sketchy. For example, the exact time of death was not recorded and, in many cases, it is not clear whether the victim was dead on arrival at the hospital or survived for a few hours. Also, the causes of death on many certificates are laconic: "Gunshot wound (riot)" with no details on the number and location of wounds. Such lapses of overworked and harried physicians, overwhelmed by the influx of several hundred wounded in addition to the dead, is understandable. It is interesting that the doctors provided more detailed information on the certificates of those who died under their care a few days after the riot than those who were dead on arrival or succumbed a few hours later.

To compound the problem, many death certificates were signed not by physicians but by Tulsa County Attorney W.D. Seavers. This was legal because at the time, state law allowed officers of the court to certify deaths that had not been attended by a physician. As nearly the entire Tulsa medical establishment was tied up in the care of the wounded, no doctors were available to examine bodies found at the scene. Apparently, this task fell to Seavers, who signed out eighteen victims whose bodies were found in the still smouldering ruins of Greenwood, or who died after being brought to temporary detention centers where blacks were held during the first hours of the riot. It is not clear whether Seavers actually visited the scene to examine the bodies or whether the death certificates were brought to him by undertakers.

Mortuary Records

At the time of the riot, the bodies of the known victims were taken from the hospitals where they were pronounced dead or, sometimes, directly from the scene to local mortuaries. There they were prepared for burial in Tulsa or shipped to other cities designated by their next-of- kin. The records of these establishments (Mobray's, Mitchell-Fleming, and Stanley-McCune), provide data on the deceased not found on the death certificates.

Press Accounts

The events of the riot received heavy coverage in local, state, and national newspapers as well as other journals, both white and black, of the time. As with all such news events, press attention was most intensive in the days immediately following the riot, then dwindled rapidly in the weeks that followed. Over the years, however, occasional newspaper feature stories and magazine articles dealing with the riot and its aftermath have appeared. The most valuable single source for these materials was the extremely thorough newspaper clippings collection from the Tuskegee Institute microfilm files.


While the white riot dead appear to have all been given proper burials, little effort was made by the white authorities to identify the bodies of black riot victims. Indeed, as both long-forgotten funeral home records and death certificates would confirm, some unidentified African-American riot victims were hurriedly buried in unmarked graves at Oaklawn Cemetery (Courtesy Greenwood Cultural Center).


Books and Monographs

Over the years, several books have been published dealing with the Tulsa race riot. These include one by a riot survivor and several others by historians who have collected written and oral accounts from survivors and their descendants.

Miscellaneous Sources

In the course of this investigation, several researchers have generously provided unpublished reports and documents on the riot which they have collected in their own studies of the event. DATA ANALYSIS

To date, death certificates on thirty-nine victims have been found. They are listed in Table 1 which summarizes the principal variables presently available on them. It should be noted that not all of the tabulated information was abstracted from the death certificates alone. For example, most of the information on the location of their wounds was found in other riot-related documents, particularly contemporary press accounts, which often provide more specific information on the nature of their injuries than was noted on the death certificates.

Table 1 Tulsa Race Riot Deaths


All thirty-nine victims, including the stillborn infant, were diagnosed as males. However, it should be pointed out that the bodies of four blacks -- all signed out by County Attorney Seavers -- were so badly burned that identification was impossible. Since it is often impossible to determine the sex in such cases without an autopsy, the reliability of a layman's diagnosis in these four cases is questionable.


Twenty-six (66%) of the thirty-nine victims, including the stillborn, were diagnosed as blacks. Again, the four bodies that were so badly burned that the could not be identified (see above) must be considered. This is especially true since thermal damage often results in the destruction of the delicate, paper-thin epidennis that is made up cells which, in blacks, contain the melanin pigments determining skin color. When this layer is extensively destroyed, it exposes the underlying dermis that, in all races, is no darker than the skin of a light complexion white person, making it easy for an inexperienced observer to mistakenly diagnosis a burned black body as white. However, in the present case, since all the burn victims were found in fire-destroyed Greenwood, it is likely that they were indeed those of blacks.


As noted above, among the black victims was an infant diagnosed as a stillborn. This case is interesting since it is apparently related to an account given to Eddie Faye Gates by a riot survivor, Rosa Davis Skinner. According to Mrs. Skinner, she and her husband Thomas, alarmed by the shooting, fled their home at 519 West Latimer a little after midnight on the night of the riot.

"When we got to Greenwood, we met up with a lot more black people who were running tying to find a safe place. We ran into a couple -- the man was one of [her husband's] best friends. The wife had just had a baby that had died at birth. She had put it in a shoe box and was waiting until morning to bury it when the riot broke out. Well durin' all that runnin' and pushin' and shovin' when black people were trying to get safely away from the riot, that po' little baby got lost! Everybody was just runnin' and bumpin' into each other. They never did find that child."

According to information in the Stanley-McCune mortuary records, sometime on June 1, police brought in the body of a newborn infant. It had been found in Greenwood earlier in the day by two white men who turned it over to the police. The body was described as that of a black male measuring "less than twelve inches long." It apparently bore no signs of trauma and was signed out as a stillborn. Like many of the other black victims, it was buried in Oaklawn Cemetery. The evidence seems compelling that the baby lost by its fleeing mother and that brought to the mortuary were one and the same. This case is important for two reasons. First, the story of this tiny victim provides a poignant glimpse of the madness that prevailed on that terrible day. Second, this infant is the only one of the thirty-nine known victims that did not die of gunshot wounds and/or burns.

Ages are given on the death certificates of all thirteen of the white victims (Table 2). One of these was apparently an estimate based on examination of the body. The others were provided by informants who knew the actual age of the victim. In contrast, ages are given for only fifteen (58%) of the twenty-six blacks and, of these, at least seven are given as estimates (usually to the nearest fifth year, e.g., "35", "40", etc.). This distribution again clearly shows that black victims were signed out with less care and regard than whites; little or no effort was made to identify blacks by contacting their next-of-kin.

Table 2 Distribution of Known, Estimated and Unknown Ages by Race

Despite the fact that no age estimates were given for nearly half of the black victims, statistical comparison of the available age data on the races is interesting. In the analysis below, I have excluded the stillborn which, as a non-violent death, is clearly a special case (see above). The mean age of white victims was around twenty-seven years compared to thirty-four years for blacks. This difference is statistically significant (Table 3).

Table 3 Age of Confirmed Riot Deaths by Race

Birthplace / Residence

The distribution of the known victims by state of birth or residence is shown in Table 4. The state of residence was inferred from mortuary records which show the state where the body was shipped for burial. This information is available in the records of only two (8%) of the twenty-five black victims. Again, an indication of the lack of attention given them before their hasty burials. This is in contrast to the whites for which birthplaces/residence of all thirteen were given. It is of interest to note that eleven (85%) of the white victims were from outside Oklahoma. The significance of this finding will be discussed more fully below. In all, natives or residents of ten states are represented among the white victims.

Table 4 Distribution of Confirmed Deaths by Race and State of Birth or Residence


Marital Status

Of the white victims, nine (69%) were single, separated or divorced. Only three were married and the wife of at least one of these does not appear to have been living in Tulsa at the time of his death. The marital status of one is unknown.

Among blacks, the marital status of seventeen is not given. Of the remaining eight, five were married and three were single.

Table 5 Distribution of Confirmed Deaths by Race and Marital Status



The occupations of ten (40%) of the black victims are known. Among them were two professionals, a physician, and a realtor (who also was a tailor). The remaining eight included five listed as "laborers", a bank porter, an iceman, and an elevator operator.

Among the twelve (92%) of the white victims whose occupations are known, there was a high school student, two cooks, a salesman, a hotel clerk, and a day laborer. Five were skilled blue collar workers and, of these, three were oil field workers; the other-two, a boiler maker and a machinist might also have been employed in petroleum-related jobs. The sole professional among the whites was the office manager of a large local oil company. Thus, at least one-third and possibly as many as one-half of the white victims were petroleum industry workers.

Table 6 Distribution of Confirmed Deaths by Race and Occupation


Cause of and Manner of Death

All of the thirteen whites were killed by gunshot wounds. Among the twenty-five black adults, at least twenty-one (84%) died of gunshot wounds. The cause of death of the remaining four, all signed out by County Attorney Seavers, were given as burn but, as noted previously, any underlying fatal gunshot wounds may not have been apparent in the absence of autopsy.

Of the thirty-nine confirmed deaths, the manner of death of all but that of the stillborn black male were homicides. The latter is classified as "natural." At least one, and possibly two, whites were killed by persons of their own race who apparently mistook them for blacks.

Table 7 Cause and Manner of Death of Confirmed Death Victims



Of the twenty-five blacks who died of gunshot injury, the wound locations of only four are documented; all four of these men died in hospitals on June 2, or later. The wound locations of the remaining twenty-one blacks, all of whom died during the first twelve hours of the riot, were unspecified. The wounds of the twelve whites whose locations are known were nearly evenly distributed by anatomical region. The overall pattern of wound distribution is rather typical of those seen in hotly contested armed confrontations carried on at moderate to distant ranges. In this, it contrasts strongly with patterns observed in extra-judicial executions by firing squads.4

Table 8 Anatomical Distribution of Gunshot Wounds of Confirmed Death Victims


Place of Death

At the time of the riot, Tulsa had four major white hospitals. Tulsa blacks were served only by Frissell Memorial Hospital, that was burned during the riot. Greenwood blacks who did not flee Tulsa altogether were first taken to temporary detention centers set up in the armory and Convention Center in downtown Tulsa. The lightly wounded who were forced to walk to the detention centers. Those more seriously injured were either carried to the centers by the unwounded or transported there by various means, including privately owned trucks and automobiles, some of which were driven by white volunteers.5

While it appears that small first aid stations were set up at the detention centers early on June 1, it must have become quickly apparent that they were not sufficient to provide the care that the dozens of wounded required. Accordingly, the basement of Morningside Hospital was hastily converted to accommodate blacks. Apparently, this makeshift facility included not only cots for the wounded but a small operating room where all surgery on the admitted blacks was performed. For the next few days, all injured blacks were treated in the Morningside basement, that may not have exceeded 5,000-square-feet of floor space.6 A brief glimpse of conditions there can be gained from a story in the Tulsa World on June 2, that noted sixty-three wounded blacks were being treated there. So far as is presently known, none of the other white hospitals in Tulsa opened their door to African American patients.

All thirteen of the white fatalities were taken from the scene to one of four hospital where they were either pronounced dead on arrival (DOA) or died later. Unfortunately, the death certificates are not always clear as to whether the victims who were admitted late on May 31, or in the early morning hours of June 1, were actually dead when brought to the hospital, or died shortly afterwards. So far as can be presently determined, at least two and possibly four whites were actually dead on arrival. All four were pronounced dead at Oklahoma Hospital by the same physician, Dr. Lyle Archerloss.

Only eight (31%) of the twenty-six black fatalities were brought to hospitals. Six died in Morningside, that as mentioned above, was the only one where blacks were treated in the first few days of the riot. A seventh died in Cinnabar Hospital on June 7, about a week after the riot. Presumably, he had been transferred from Morningside after Cinnabar had been reopened. The last died on August 20, in the Red Cross hospital that was set up in the Greenwood's black Dunbar School after the riot.

The other eighteen (69%) blacks were not taken to hospitals. The bodies of these sixteen individuals were found in the downtown area where the fighting began or in the ruins of Greenwood. Five days after the riot on June 6, the badly decomposed body of a black man was found about eight miles east of Tulsa. He had died of a gunshot wound of the neck. He was later identified as a man who had escaped from a temporary detention center.

All of these bodies were taken directly to mortuaries and their death certificates were signed out by County Attorney Seavers. Another of these "non-hospital" victims died in the armory detention center where he was taken after he was shot down by a teen-aged member of the mob while trying to surrender outside his home in Greenwood. Ironically, this man -- a prominent physician -- lay without medical attention for several hours before he finally succumbed to a bullet wound of the chest. His death certificate was also signed by the county attorney.7

Table 9 Distribution of Confirmed Deaths by Place of Death


Date of Death

The records indicate that four of the white casualties died before midnight on May 31. If this is correct then these men were most likely killed in the downtown area where the fighting first began. Seven others died on June 1 and one on June 2. The last white fatality died in the early morning hours of June 6. He was wounded a few hours earlier when white militia men fired on the car in which he was riding. The perpetrators, a least one of whom was wearing his World War I army uniform, claimed that the driver of the car refused to obey their orders to stop.

None of the twenty-six black victims is listed as having died on the evening of May 31. Twenty-one were signed out as having died on June l, two on June 2, and two others on June 7, and June 10, respectively. The last black to die of riot wounds was a twenty-one year old who lingered until August 20, eleven weeks after the riot.

The fact that no black fatalities were recorded for the evening of May 31, is curious. According to several sources, many shots were fired by both sides during the retreat of the blacks from the courthouse area back to Greenwood, and some early newspaper accounts describe blacks lying wounded or dead in the downtown area. If the latter are true, it suggests that no medical aid was extended to those wounded blacks unfortunate enough to have been left behind during the retreat to Greenwood.

Table 10 Confirmed Deaths by Date of Death



As in most of the United States at the time, Tulsa mortuaries were racially restricted. The three major establishments serving white Tulsans were Mitchell-Fleming, Mowbray, and Stanley-McCune. Black funerals were handled by a single Greenwood funeral home operated by S. M. Jackson, a graduate of the Cincinnati (Ohio) School of Embalming. In 1971, Jackson was interviewed by Tulsa historian Ruth Avery.8 His account of his riot experiences is valuable since it provides some insight into the way the dead, both black and white, were handled. On the morning of June 1, when the white mob stormed into Greenwood, Jackson's funeral parlor was burned down. At the time, he was holding four embalmed bodies for burial; only two of these were retrieved (leaving one to wonder about the fate of the other two). At first interned, he was promptly paroled by the owners of Stanley-McCune who temporarily hired him to help process the bodies who were brought to their establishment. During the next few days he embalmed several blacks whose bodies were to be shipped to other cities for burial (see below).

Stanley McCune also had a hastily arranged contract with Tulsa County to bury (unembalmed) the bodies of blacks whose relatives could either not afford to claim them for private burial or were not informed of the deaths. In all, Stanley-McCune handled the arrangements for two whites and eighteen blacks. The bodies of all of the blacks were prepared for burial by Mr. Jackson. He embalmed two of these that were claimed and were buried in other cities. The remaining sixteen were not embalmed and placed in plain wood coffins. Mr. Jackson was able to rebuild his Greenwood business and handled the funeral of the last black riot victim who died on August 20 and whose body was claimed by his family for burial in his native Mississippi.

Table 11 Distribution of Confirmed Dead by Mortuary


Burial Places

Only three of the white victims were buried in Rose Hill, a privately operated cemetery. Another was buried in Watonga, a small town in western Oklahoma. The remaining nine were buried in other states. Five of the black fatalities were buried outside of Tulsa: two in other Oklahoma towns and three outside the state. The remaining twenty-one blacks (84%) were interred in Oaklawn, the Tulsa municipal cemetery.

Table 12 Burial Places of Confirmed Dead


The Oaklawn Burials

In light of the controversy surrounding the total number of black victims of the race riot and the disposal of their bodies, the documented burials in Oaklawn take on a special significance. This is especially true in the light of the preliminary archaeological findings.9

As noted above, twenty-one black victims, 84% of the total, were buried in Oaklawn. At that time, the cemetery was segregated by race and blacks were buried in the western-most section, so it is safe to assume that these black riot victims also were buried there. Five of these victims, all of whom died in Morningside Hospital, were buried by Mowbray mortuary. All these hospital cases died of gunshot wounds. Their death certificates were signed by a single physician, J. F. Capps, M.D. Dr. Capps signed out two of these as "John Does." Four died on June 1, and the fifth in the early morning of June 2.

The remaining sixteen were bodies found at the scene and taken to Stanley-McCune; their death certificates were signed by County Attorney Seavers. Six of these, four of whom were badly burned, were not identified. A seventh unidentified body was that of the previously described stillborn. The remaining nine were identified.

These Oaklawn burials were conducted at county expense. The Mowbray and Stanley- McCune records indicate that the victims were not embalmed but buried in plain wooden coffins; they also show that the mortuaries charged the county $25 for each burial. An important feature of the Stanley-McCune records was a notation indicating the "grave number" of each burial. These numbers form a single sequence from 1 to 19, except for graves 15, 16 and 17. It is possible that these graves were filled by three of the Mowbray. Unfortunately, grave numbers were not given in the Mowbray records.

The data currently available on these Oaklawn burials is given in Table 13. They are significant for several reasons. First, should archaeological exploration of the area go forward, the excavators should encounter them. Assuming, as the records indicate, that they were buried in separate graves in the order indicated by the Stanley-McCune grave numbers, they should be encountered in an orderly row(s). If so, the available information that we have on them should be valuable in obtaining tentative identifications. For example, the skeletons in graves 7, 9, 13, and 18 should show some signs of fire exposure. If so, they should provide tentative leads to the non- burned skeletons in adjacent graves.10 By narrowing the number of possible decedents, the effort (and the cost) of DNA identification could be substantially reduced.

Table 13 Burials of Confirmed Dead in Oaklawn Cemetery



Of course, this small group of documented fatalities cannot be considered a statistically- defined random sample of those who had some role in the riot, either as active members of the mob or as passive victims. However, it is probably typical enough to provide some glimpses of the kinds of people who were caught up in the riot.

The whites ranged in age from sixteen to thirty-nine years. As a group, they tended to be young, with a median age of twenty-seven years. The state of birth or residence of all thirteen are known and, of these, only two were born in Oklahoma. The bodies of all but four were shipped to other states for burial and, of the four Oklahoma burials, only three took place in Tulsa. Of the ten for whom we have marital information, seven were single, one was divorced and another had been separated from his wife for nearly twelve years. Among the three married men, the wife of one was not living in Tulsa at the time of the riot. At least four and possibly six were employed in petroleum-related jobs; three others held jobs suggesting transient status: two were cooks and the third listed as a "laborer." Judging from their occupations, all were of lower socioeconomic status except one, an oil company junior executive.

In short, the limited demographic information that can be drawn from such a small sample indicates that these men were probably fairly typical of white Tulsans of the oil boom days: young, single, non-professionals from outside Oklahoma who had been lured to Tulsa by the promise of good jobs and good money. With no strong domestic ties to keep them home that night, drifting around in the bustling downtown area on a nice summer evening, perhaps looking for ladies, liquor or other excitement, they also were the kind who might be expected to show up around the courthouse when the talk about lynching a black accused of assaulting a white girl got started. Since boot-legging was a busy cottage industry in Tulsa, it is possible that at least some of them had high blood-alcohol levels by the time the trouble began.

Black victims, in contrast, tended to be older than whites. They ranged in age from nineteen to sixty-three. Blacks averaged close to 35 years in age -- nearly seven years older than the whites. This difference is statistically significant. Of the eight for whom marital data is available, five were listed as married. While their occupational status tended to be lower than that of the whites (and none were employed in the petroleum industry), two, a realtor who also owned a tailor shop and a highly-regarded physician, were solidly middle class. Unlike the whites, most of whom were young, single, newcomers to Tulsa, this group of black victims appears to have been stable, older citizens of the Greenwood community.

These thirty-nine cases also demonstrate that, compared to white victims, those who were black victims were treated with what would today be considered cavalier, if not criminal, carelessness. This is indicated by the fact that at least one was allowed to bleed to death without medical attention in a detention center instead of being taken immediately and directly to a hospital after being gunned down in Greenwood while trying to surrender. Another indication of this is found in the death certificates. Those of at least four of the thirteen whites were pronounced dead before midnight on May 31, indicating that they were promptly taken to hospitals. In contrast, none of the death certificates of black victims are dated earlier than June 1, a finding that suggests that whether dead or still alive, they lay unattended for at least several hours. More evidence is provided by the fact that adequate treatment facilities were denied blacks until sometime in the late morning or afternoon of June 1, when a makeshift ward and surgery was hastily set up in the basement of one of the several hospitals that normally admitted only whites. Only then were the many black wounded provided with care, and some allowed to die under the care of nurses and physicians.

If Tulsa medical care givers were callous and careless in their treatment of black riot victims, representatives of the Tulsa funeral industry were not far behind them. This is shown by the hasty, "county" burials in Oaklawn on June 1 and 2. Their death certificates in most cases signed by a layman, County Attorney Seavers. Much of the vital information on these certificates such as address, age, marital status, next-of-kin, etc. was left blank or filled in with a hastily scrawled "don't known". This indicates that authorities with the responsibility to contact families and identify victims did not bother to track them down in the admittedly crowded and confused detention centers. Thus, some families that might have been able and willing to claim their dead and bury them properly were not given this opportunity. Whether they could afford to or not, most probably did not know for sure that their relatives were already dead and buried in unmarked pauper graves until they were released from detention.

Another finger of blame points to law enforcement authorities at the local and county levels. As noted previously, all of these deaths -- both black and white -- were homicides which occurred within the jurisdiction of either the Tulsa Police Department (thirty-seven cases) or the Tulsa County Sheriffs Department (two cases). Yet, so far as is known, these murder cases were not investigated while at least some of the perpetrators could be identified and apprehended. Prosecutorial authorities, both county and state, also are accountable since they apparently did not aggressively press for such investigations.

These hard truths cannot be presented without pointing out that many white Tulsans and Tulsa institutions (particularly some churches and the local Red Cross) took a courageous role in the riot by offering protection and care to their black neighbors. Their brave actions have been well documented elsewhere and will not be considered in detail here.

It should also be pointed out that what happened in Tulsa could have taken place in almost any other city in the United States in 1921. Nor were the conditions and circumstances leading to this tragic event a uniquely Oklahoman, or even "Southern" phenomenon. In the data considered here, this is probably best illustrated by the known birthplaces or residences of the white fatalities. Of the thirteen men who were killed, only two were native Oklahomans. None were from states of the deep South. Five -- the two Oklahomans, a Texan, an Arkansan and a native of Kentucky -- were from Confederate border states in which the populations were of deeply divided loyalties during the Civil War. The remaining seven were from midwestern or northeastern states.


In summary, perhaps the least that can be said of the physicians, undertakers, police, and prosecutors of Tulsa of the time was that they were not hypocritical: they treated their black fellow-citizens no better when they were dead than they did when they were alive.

Although this preliminary report is limited to treatment of the confirmed dead, it cannot be closed without considering the as yet unconfirmed dead of the Tulsa race riot. First to be considered are the eighteen deaths that occurred in the Maurice Willows Hospital operated by the Red Cross until January 1, 1922. A systematic search of vital statistics records to find their names and the causes of their deaths has not yet been made. Some may have died of complications of wounds received during the riot; if so, of course, such deaths would add to the riot deaths. Others, particularly, if children or elderly whose homes were destroyed or their family life disrupted, may have succumbed easily to diseases they may have otherwise survived; while actually not killed in the riot the deaths of these victims would certainly have to be considered as riot-related.

As noted in the introduction of this preliminary report, we already have the names of many possible decedents and, hopefully, may obtain still more. These reported dead will first be scanned against vital statistics records to see if their death certificates have been somehow overlooked. If they are not found, it will not necessarily mean that they did not die in the riot since there is at least some tenuous evidence that more people, especially blacks, died in the riot whose deaths were not recorded. Most of this evidence, it is true, is in the form of wildly varying estimates that appeared in both the Tulsa and national press in the days and weeks immediately following the riot. Many Tulsans, white and black, have recollections of bodies of victims being disposed of in irregular ways in the first few days following the riot. These estimates and stories cannot be dismissed lightly.

As one whose entire professional life has been devoted to the investigation of mass disasters such as fires and floods, aircraft accidents, human rights violations, war crimes and acts of terrorism throughout the world, this writer is fully aware of the often exaggerated estimates of the number of victims that surface in the wake of the chaos and confusion following such events. At the same time, experience has shown that in manner of these situations, official counts of the dead or often seriously underestimated.

In the present case, it should be pointed out that, like nearly all other states at the time of the riot, Oklahoma had no adequate system for the medicolegal examination of violent or unattended deaths. Today, the law mandates that all such deaths fall within the medicolegal responsibility of the State Medical Examiner. Bodies of such victims are examined and, when necessary, autopsied by forensic pathologists to determine the cause and manner of death. At the time of the riot, the law required that death certificates be signed by attending physicians or, as we have seen, certain public officials in exceptional cases. However, it appears that there was no controlling legal authority (to use a phrase currently in vogue) that required that medically unattended deaths not coming to the attention of officers of the court be documented with a state death certificate. Therefore, it is possible that bodies found in the ruins of Greenwood during the days immediately after the riot were simply buried without documentation.

That this may have indeed happened is suggested by a statement apparently made by Major O. T. Johnson, a Salvation Army officer stationed in Tulsa at the time. According to stories in at least two newspapers, the Chicago Defender, June 11, 1921 and St. Louis Argus, June 10, 1921, Johnson is said to have stated that he hired a crew of over three dozen grave diggers who labored for several days to dig about 150 graves for Negro victims. Unfortunately, any official report that Major Johnson may have submitted to the Salvation Army has not yet been located. However, the possibility the statement attributed to him was indeed true is at least partly supported by two witnesses. One, Eunice Cloman Jackson, the wife of black mortician S. M. Jackson stated in 1971 that her step-father was part of a crew of fifty-five grave diggers; when she was asked where the bodies were buried, she replied that ". . .most of them were out at Oaklawn. That was the cemetery for burying them. . .."11 Clyde Eddy, a young boy at the time, remembers seeing large wooden crates, each containing several burned bodies, awaiting burial in Oaklawn in the days following the riot. If bodies were collected from the burned out area of Greenwood they may well have been collected in crates rather than individual coffins and transported to Oaklawn for burial by Major Johnson and his large crew of grave diggers. They most likely wood have been carried on trucks, railroad flatcars (the Frisco tracks ran adjacent to Oaklawn), or both, thus accounting for the several eyewitness reports that bodies were seen being carried from the Greenwood area on both trucks and flatcars.

The theory that perhaps as many as 150 bodies were buried in Oaklawn under Major Johnson's supervision can be framed as an hypothesis that can be tested by archaeological exploration of the area described elsewhere in this volume by Drs. Brooks and Witten.12 Such an effort would, at the least, result in the recovery of the twenty-one black confirmed dead from their unmarked graves so that they can be more suitably memorialized and, possibly, identified. If the hypothesis turns out to be true, it would result in the recovery of the bones of the undocumented dead and, thus, help provide a solution to a lingering mystery.






1Theoretical indeed, since at this late date the perpetrator most likely would be as dead as his victim and the case, thereby, moved to a higher (or, possibly, lower) jurisdiction.

2The geriatric problems of conducting such a trial would be a nightmare. Imagine the complications resulting from the inter-tangling of iv and catheter tubes of the witnesses and defendant as they traded places on the witness stand!

3Tulsa World, June 3, 1921.

2The geriatric problems of conducting such a trial would be a nightmare. Imagine the complications resulting from the inter-tangling of IV and catheter tubes of the witnesses and defendant as they traded places on the witness stand!

3Tulsa World, June 3, 1921.

4Snow, Clyde. 1993 Forensic Anthropology Report. in Anderson, Snow et al. The Anfal Campaign in Iraqi Kurdistan: The Destruction of Koreme. (Middle East Watch/Physicians for Human Rights, New York and Boston: 1993).

5At this time, the three or four ambulances in Tulsa were operated by mortuaries and it appears that all of them were fully employed in taking wounded whites to the various hospitals.

6Warner, personal communication, November 11, 2000.

7What a excruciatingly cruel fate for a physician to have his death certificate signed by a lawyer!

8Avery, R. "African-American S.M. Jackson (Mortician) and his wife, Eunice Cloman Jackson on June 26, 1971 ", unpublished transcript of taped interview.

9See the report of Drs. Brooks and Witten elsewhere in this publication.

11Eddy, loc. cit.

12 Brooks and Witten, loc. cit.





 Table Of Contents -- Home