A Qualitative as well as Quantitative Database
Our goal is to create a collaborative database on the history of violent crime, violent death, and collective violence from medieval times to the present. Historians, social scientists, and genealogists must work together if we are to gather enough data to allow researchers to describe accurately the history of violence. No single researcher or group of researchers can examine enough sources in enough jurisdictions to achieve that goal. We are confident, however, that an ongoing, collaborative effort can.
We envision the Historical Violence Database as a complement to the computerized databases on crime and mortality that have been maintained by the World Health Organization since 1950, and by the United Nations, Federal Bureau of Investigation, and the National Institutes of Health since the early 1970s, and to the database that will be created by the new National Violent Death Reporting System at the National Center for Injury Prevention and Control in Atlanta, Georgia. In future years, the NVDRS will gather up to 200 details on each violent death in the United States (including, for instance, for each case of homicide, the length of the gun barrel if a gun was used, the level of education of the victim and perpetrator, whether drugs or alcohol were involved, etc.). Our historical database cannot be as comprehensive because of gaps in the historical record and because of omissions (sometimes deliberate) by record keepers. The historical record is more complete, however, than many people realize. People in the medieval and early modern periods, like people today, were deeply concerned about violent crimes, violent deaths, and collective violence. They kept abundant, if incomplete records of each. In many instances, statistical methods can estimate the number of cases that are missing from the surviving records (Eckberg 1995, 2001; Monkonnen 2001a). And the surviving records can be supplemented by the findings of forensic archaeologists, who can estimate the incidence of certain kinds of traumatic injuries from skeletal remains (Larsen 1997; Walker 1997, forthcoming; Roth 1999, 2001a). Our hope is that historical research, statistical analysis, and forensic science will enable researchers to trace the history even of such elusive crimes as child murder and spouse abuse.
For a scholarly introduction to the database, please see:
Randolph Roth, Douglas L. Eckberg, Cornelia Hughes Dayton, Kenneth Wheeler, James Watkinson, Robb Haberman, and James M. Denham. 2008. “The Historical Violence Database: A Collaborative Research Project on the History of Violent Crime and Violent Death.” Historical Methods 41:81-97.
The purpose of the Historical Violence Database is to help scholars better understand violent deaths (homicides, suicides, accidents, casualties of war), serious assaults (attempted murders, sexual assaults, arsons, maimings, aggravated assaults), and contentious political action that leads to violence. The database will allow social scientists to test their theories of violent crime, violent death, and collective violence in a variety of historical circumstances. Many theories concerning, for instance, the relationship between capital punishment and homicide rates, or the relationship between guns laws and armed robberies, fail once they are forced to confront data from 1930s and 1940s, rather than the 1980s and 1990s. The database will also allow historians to share their research so that others can build upon it. Many historians who have written on violence (usually on the state or county level) have failed to preserve their notes or share their research. Thus, their evidence cannot be checked for accuracy or reanalyzed. We hope that the Historical Violence Database will help social scientists transcend the limits of contemporary data and historians the limits of irreproducible local studies, so that we can better describe and explain the history of violent crime, violent death, and collective violence.
A Qualitative as well as Quantitative Database
The Historical Violence Database gathers qualitative as well as quantitative data on each suspected act of violence that appears in the historical record. Most databases on crime, death, or collective violence are quantitative. They report the bare facts: date, place, cause of death, etc. Such databases seldom report qualitative information: the details of the incident, the personal histories of victims or perpetrators, the ambiguities in the surviving evidence, etc. Without such information, scholars cannot reclassify crimes or deaths or protests or ask new questions about them. Was the cause of death clear? Did every witness tell the same story about the incident? Was the guilt of a person convicted of a particular crime in doubt? Would a victim have survived with the benefit of modern medicine? These questions cannot be answered unless quantitative information is supplemented by qualitative information.
As important, existing databases on crime, death, or collective violence seldom report the kinds of qualitative information that historians need to understand events in their social and cultural context or to place events in their proper sequence. What words or gestures did victims or perpetrators use in violent encounters? How did bystanders react? Did such encounters follow scripts? Were there common threads in the life stories of certain kinds of suicide victims? What were the material or emotional circumstances of their lives? Historians cannot answer such questions without qualitative information. They need qualitative information to study individual cases in detail.
That is why the database will include five types of files: worksheets, spreadsheets, images, text files of original documents, and relational files.
The worksheet files contain a record of each researcher's notes on each case, the sources consulted, and the interpretive decisions made. The worksheets will contain all the information on each case, qualitative as well as quantitative. The worksheets will give future researchers an opportunity to check references, correct errors, add information from new sources, and reinterpret individual cases or groups of cases. These files will be maintained in Microsoft Word. Handwritten notes of researchers, if available, will be maintained in Abode Acrobat files (pdf). Codebooks, templates, and examples of worksheets.
The spreadsheet files contain a record of the quantitative information on each case. They are at present victim-based. The spreadsheet files will allow future researchers to manipulate the data statistically and discover new patterns in the data. These files will be maintained in Microsoft Excel and in Comma-Separated-Variable (CSV) files (although several are as of yet available only as SPSS files). Examples of spreadsheets.
The image files contain photographic images of original documents on violence: coroner’s reports, homicide detective reports, newspaper articles, local histories, etc. They offer historians an opportunity to examine the evidence first-hand and make their own judgments. Most of these images will be maintained as JPEG or Adobe Acrobat files. Examples of image files.
Text Files of Original Documents
Printed original documents on violence will be maintained in Microsoft Word if they are in the public domain and optical character recognition (OCR) is possible. Examples of text files of original documents.
The relational files (which have not yet been created) will organize the quantitative information on violence in a different format. They will contain information on each victim and assailant, and will link them to a particular crime or crimes. Relational files are more flexible than victim-based spreadsheet files, because they allow researchers to study victims, assailants, or relationships between victims and assailants. Relational files can link multiple victims to a single perpetrator or multiple perpetrators to a single victim. These files will be maintained in Microsoft Access.
A Living Database
Our plan is to improve the Historical Violence Database over time. We will add data from new jurisdictions and new sources as they become available. We will change our worksheets, codebooks, and file formats as we learn from experience and from the recommendations of readers and contributors. And we will revise the data as new sources become available and as errors come to light. We encourage researchers to notify us of any mistakes or of any new cases or sources that they find. We will gladly acknowledge their contributions in the database.
We plan to take advantage of scanning and photographic technology as it improves, so that researchers can read entire newspaper articles or inquests on particular cases, rather than researchers' notes on such articles. We also hope to scan any handwritten worksheets on violent crimes or deaths that researchers would like to contribute, whatever their format, so that they can be made available to other scholars. We would also like to archive any computerized historical data that researchers would like to share with the project or establish links to the websites where their data are available. The fundamental goal of the database is to encourage the sharing and preservation of data, not to impose a single format for data collection.
We also hope to create a comment page where researchers can debate the interpretation of individual cases or groups of cases. Researchers will also be able to post questions concerning sources, sources, theory, etc.
Our goal is to consult multiple sources wherever more than a single source is available on a particular act of violence. Statistical procedures for estimating the number of violent incidents from fragmentary records require "matching lists" from two or more independent (or largely independent) sources, such as court records and newspapers (Eckberg 2001, Monkonnen 2001a, Roth 2001a). The procedures do not require that the lists be perfect—only that they be compiled from different sources. That is why the project does not rely solely on court records or vital records.
We are interested in gathering any kinds of data relevant to the study of violent crimes, violent deaths, or collective violence. For example, we are studying gun ownership, so that we can search for possible relationships between weapons and violent crime, and we are studying abortions and abortion-related deaths, so that we can search for possible relationships among fertility, contraception, illegitimacy, abortion, and violence against children. We are interested in data on all possible correlates of violence.
Scope of the Project to Date
To date, the project has focused on the history of violence in Western Europe and the United States, especially New England. Randy Roth, Nina Dayton, and Robb Haberman have worked through a variety of sources for New England (except Rhode Island) from 1630 through the Revolution. Randy Roth has done the same for Vermont and New Hampshire through 1900. We have studied court records, docket books, case files, inquests, government publications, newspapers, tracts, diaries, town histories, vital records, cemetery records, and genealogies, among other sources. The sources are discussed in Roth (2001a: 126-9) and Dayton (1995).
Ken Wheeler has studied homicides in Ross County, Ohio, 1798-1900, and in Holmes County, Ohio, 1825-1900 (Wheeler 1993, 1997). Randy Roth and Ken Wheeler have gathered data on violent crimes in five counties in northern Georgia, 1779-1900 (Franklin, Gilmer, Jasper, Rabun, and Wilkes). Randy Roth and Jim Watkinson have gathered data on violent crimes and inquests in eleven counties in Virginia, 1634-1800 (Amelia, Augusta, Botetourt, Lancaster, Middlesex, Richmond, Rockbridge, Rockingham, Spotsylvania, Surry, and Sussex) and in four Virginia counties, 1800-1900 (Amelia, Lancaster, Rockbridge, and Surry). James Campbell has gathered data on violent felonies and misdemeanors in Richmond, Virginia, 1830-1861, and James Rice on crime and criminal justice in Frederick County, Maryland, 1748-1837. Phil Schwarz has studied slave crimes (including violent crimes) in numerous counties in Virginia, 1705-1865. Mike Denham has studied crimes in Florida, 1821-1861, and Doug Eckberg is studying homicides in post-Reconstruction South Carolina. Glenn McNair is studying slave crimes in Georgia, 1751-1865; John Coombs is studying crime and violence in colonial Virginia; and Terri Snyder is studying suicides in Virginia, 1607-1830. Gilles Vandal is studying crime and violence in Louisiana, 1840-1885. Eric Monkkonen completed studies of homicide in New York City, 1790-1999, and in Los Angeles, 1830-2002. Petula Iu is continuing Eric’s work on Los Angeles homicides. Leigh Bienen has created a database on homicides in Chicago, 1870-1930, from the homicide reports of the Chicago Police Department. Kevin Mullen has created a similar database on homicides in San Francisco, 1849-2000, using newspaper accounts and the homicide reports of the San Francisco Police Department. Bud McKanna has gathered data on homicides in a number of counties in the Trans-Mississippi West, 1850-1920, and Jack Marietta and Gail Rowe have done the same for all violent crimes in colonial and revolutionary Pennsylvania. Paul Gilje has contributed his data on over 4,000 riots in America from colonial times to the present. Roger Lane has retired from writing history, but he has donated his worksheets on homicides in Philadelphia, 1839-1932. Michael Maltz has just completed a revision of the Uniform Crime Reports of the Federal Bureau of Investigation, 1960-2004.
Barbara Hanawalt has studied crime and violent death in thirteenth- and fourteenth-century England. Carolyn Conley has investigated homicide in Ireland, Scotland, England, and Wales, 1867-1892, and Pieter Spierenburg has studied homicide in Amsterdam from the 1400s to 1800. Mary Beth Emmerichs is studying homicide in nineteenth-century London. Howard Brown has studied crime and criminal justice in revolutionary France; Richard McMahon is studying homicide in nineteenth century Ireland; and Andrew Stickley is studying homicide in European Russia in the late nineteenth and twentieth centuries.
The data will be posted on our website and archived with the National Institute of Justice and the Inter-University Consortium for Political and Social Research, once our major publications from the project are completed. Until then, we will post samples of the data on our website.
A full scholarly discussion of the Historical Violence Database, including its methods and purpose, is available in:
Roth, Randolph, Douglas L. Eckberg, Cornelia Hughes Dayton, Kenneth Wheeler, James Watkinson, Robb Haberman, and James M. Denham (2008) “The Historical Violence Database: A Collaborative Research Project on the History of Violent Crime and Violent Death.” Historical Methods 41: 81-98.
Terms on Which the Data and Text Files in Historical Violence Database May Be Used
Creative Commons Deed
The following are the terms on which the data and text files in the Historical Violence Database may be used.
You are free to copy, distribute, display, and use the data to make derivative works from the data.
Under the following conditions:
Attribution: You must give the original author credit for gathering the data.
Noncommercial: You may not use the data for commercial purposes.
Share Alike: If you alter, transform, or build upon the data, you may distribute the resulting data sets only under a license identical to this one.
For any reuse or distribution, you must make clear to others the license terms of these data. Any of these conditions can be waived if you get permission from the copyright holder. Your fair use and other rights are in no way affected by the above.
This work is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 United States License
This is a human-readable summary of the Legal Code (Read the full license on the Creative Commons website).
License type: Attribution-NonCommercial-ShareAlike 2.0