WEB construction--a collection of notes

There are numerous errors and inconsistencies—such as incorrect crime types, misspellings, mistaken identities, and cases where execution status cannot be confirmed—and that some entries have been corrected or clarified using more recent historical scholarship and archival research. And the best resource is still WEB. Occasionally I will describe ongoing work to develop variables that capture demographic information, crime categories, victim information, and contextual factors. Some variables, such as race or age, remain incomplete for early historical cases and require additional genealogical work. Crime classifications sometimes differ between Espy’s records and Baker’s interpretations, and the project aims to reconcile these differences by consulting primary sources whenever possible. Victim information is also being expanded so that infanticide cases and other forms of homicide can be more accurately understood. The Espy index cards are a starting point, supplemented by archival materials, court records, compensation claims, and established historical research. Every variable is manually coded, with attention to transparency about uncertainties or conflicting sources.

A Selection of dataset notes pertaining to changes/additions (11/24)

The earliest woman recorded as executed appears to have been sentenced for an extramarital relationship with a Black man. According to the source material, her sentencing occurred in 1630. The Espy card, however, lists her as charged with infanticide in the Virginia Colony in 1633. Multiple historians—Ann Jones (Women Who Kill, 1980) and Julia Cherry Spruill (Women’s Life and Work in the Southern Colonies, 1938)—cite Hening’s Statutes at Large (Vol. I, p. 209) and the Virginia Magazine of History and Biography (Vol. XIII, p. 390). Spruill notes that Hatch pleaded pregnancy after conviction, but a jury of matrons determined she was not with child.

A later note by researcher Hearn (Dec. 9, 1989), based on the Minutes of the Council and General Court of Colonial Virginia, offers a more nuanced view: Hatch was acquitted of murder but convicted of manslaughter. She received a death sentence only because she failed to plead clergy—a legal right that usually prevented execution. Hearn doubted she was actually executed and recommended marking the execution as unconfirmed.

One woman described in Winthrop’s Journal (History of New England, 1630–1649, Vol. 1, p. 238) was said to be mentally ill or “possessed by spirits” and claimed Satan instructed her to kill her three-year-old child. Winthrop was likely biased, but the case is notable for the suggestion that she used elements of Hutchinson’s theology to aid her defense.

Inquiries have been sent to several archives, including the South Carolina Department of Archives and History, and this portion of research is considered complete for now.

Hearn (1999, p. 10) also clarifies that V17 and V18 should be dated to 1644, not 1643. Both defendants failed the “touch test.”
Another case involved an enslaved woman possibly raped by her master (Jones, pp. 74 and 82). She was reportedly “hanged twice” because the rope failed on the first attempt.

Depression appears likely in one case involving a trial by touch; her execution date has been updated to 1648 based on Baker, though a Cotton Mather diary entry is still sought for confirmation. This case also reveals an Espy error: Mary Parsons and Mary Young were confused in the data. (See the Albany archives for details.)

In another case, a woman and her husband were both accused of witchcraft. Baker notes she did not make execution easy for the officials. Espy sometimes recorded her execution as occurring in Boston, though Suffolk is adjacent and may explain the confusion. Espy’s notes also reference rumors about her reputation (e.g., being a “harlot” or possibly committing abortion). Putnam Demos’s Entertaining Satan (1982) is cited as a valuable reference.

Mary Paine (later Parsons) was born in England, which may explain discrepancies in her reported birth year. Evidence suggests she identified two other alleged witches: Knapp and Staples.

In another entry, a book from V31 in the Espy collection confirms material where the card is missing. Several cases strongly imply execution occurred without explicit confirmation. One woman was said to have died for essentially being argumentative (Baker). Another was executed on a boat as an attempt to calm stormy seas.

A Maryland case involves a woman who served Margaret Brent and was originally Mary Lawn before marrying Dan Clocker. She received a death sentence but may not have been executed; she died the following year.

Another person appears to have been executed at sea (1654 or 1659 depending on the source). Mary Dyer, an important figure in Rhode Island’s history, was well respected, wealthy, married, and a Quaker missionary. She was close to Anne Hutchinson. Baker notes that Hutchinson’s death in 1643—traditionally blamed on an attack—may have been indirectly caused by Puritan authorities. Her daughter Susanna survived and lived with Indigenous people for four years.

Date corrections: she was executed with Rebecca Greensmith and her husband—the last witchcraft hangings in Connecticut. Execution may have been in 1661; a 1988 letter from Hearn argues convincingly that she was executed.

Possible confusion exists between Parmelia Yarber and “Parmelia” (Isabella) noted in Baker. One case was severely miscoded: the crime was arson, not infanticide.

One husband may have poisoned his first wife (Romero & Crane, 2020). Another case is part of the broader Salem witchcraft events; occupation corrected to “housewife” and county to St. Mary’s, though Baltimore is possible. Ashley Ellefson's 2009 research helps clarify some of these details.

Some cases remain marked unconfirmed pending additional verification. One involved an extramarital pregnancy with an African American man (earlyamericancrime.com was helpful). Another execution date was corrected from 1708 to 1706. One woman disemboweled her husband; others were involved in the New York slave insurrection (see note for “Unnamed,” 1708). Compensation claims appear several times, raising questions about the legal or owner-related contexts of the executions.

One may be the last public burning. If executed, the method would have been burning; V24 updated accordingly. Several women had children who died shortly after birth, sometimes involving Black fathers; some cases involve Pequot histories and early Connecticut executions.

Other corrections involve counties, alternative spellings, misattributed state data, and identification of proper victims. Some cards appear not to be in Espy’s handwriting. In one Maryland case, the Maryland Gazette notes the prisoner denied guilt until the end; she may have been pregnant in jail.

A number of women appear in the records only tangentially but are tied to major events, including war-provoking acts or highly public executions. Dent and Floyd appear on one notecard with uncertain outcomes. One woman received a gubernatorial reprieve that came too late.

Slaves were sometimes executed alongside spouses; some cases involve additional victims, transfer of bodies to medical colleges, or county-level corrections. Several cards reference compensation amounts (e.g., $350, $387.50, $400, $600). Some women were noted for repeated trials, prison escapes, or self-defense claims. Others were recorded as innocent by later review.

Some women were described as turbulent, mentally impaired, or extremely young. A few were labeled suspected serial killers. A number of entries involve pregnancy, delayed executions, or unclear racial classifications (verification needed). Espy’s file often appears to be the sole surviving documentation for some 20th-century cases. In several, boyfriends or family dynamics contributed directly to the circumstances of the crime.

Later entries (e.g., Velma Barfield, Betty Lou Beets) reference modern legal appeals. Ross (2017) discusses links between carpentry and witchcraft accusations. Additional heavily detailed cases (including those involving poisoning, child deaths, and exploitative relationships) come from Hearn, Baker, Meacham, Streib, and others. Many original index cards can be found in the Albany digital collections.

The initial identification of executed women from 1632-2023

Description of Database Construction

Importantly, for those seeking verified data for men and women in the Espy File and, as I understand it, no better datasets exist for these five states, see the Espy Project Datasets, 2015-2020. Excel spreadsheets for five states are complete! As per website "In 2016, the Council on Library and Information Resources (CLIR) funded The Digital Archive of Executions in the United States, 1608-2002: Digitization and Access to the M. Watt Espy Papers to digitize the index card summaries and reference material and make them available online. The Project attempted to digitize the index card summaries and the reference materials from the Espy Papers and unify them with the Espy File dataset created from the collection during the 1980s."

To date, there was no one electronic dataset of women and capital punishment that encompasses the many variables that affect sentencing outcomes and general treatment in the criminal justice system. And, to date, it seems the most complete listing found in one resource is still David V. Baker's (2016) book Women and Capital Punishment in the United States, to which I owe for the very framework of the dataset and the work I've since produced. As of now, the dataset the executed women as identified by the Espy File, Baker's (2016) table on pages 43-61, and the DPIC. Any new names identified by Baker were added to the Espy file (indicated in red).

The project took longer than anticipated because each variable was initially entered manually. Specifically, coding was not initially a concern because of an overestimation of the differences that existed between the Baker set and the Espy data. After completion, many similarities were observed. To facilitate comparison and given that the Capital Punishment project for the Espy project utilizes many of the same codings, relying on many of the Espy files' codings--would simplify transmissibility and comparisons between datasets. It's important to note that the correct identification of men and the legitimacy of their executions were not verified for this project. Only women executed in the United States were verified. However, the ICPSR website, frequently referenced in the code book and throughout the website, does provide an idea for those wanting to make comparisons to men.

I do apologize for the spreadsheet as producing the data are my first priority, my book is a close second, and therefore the aesthetics of it are far down the list. I would like to make the data more accessible and, should anyone inquire, it would be just the impetus I need to do so.

Data Sources and Discrepancies

Streib's 1992 article is excellent and provides a much-needed clarification as to how later estimates of women's executions were calculated and how he initially utilized the research of M. Watt Espy. The M.E. Grenander librarians and archivists created an easy guide for locating his material, much like for Espy, for the Victor L Streib Papers as housed in the National Death Penalty Archives. "Much of Streibs scholarship has been aided by Watt Espys collection of death penalty files, and Streib has stated that his work is credited to "standing on the shoulders of Watt Espy." " as quoted on the website: https://archives.albany.edu/description/catalog/apap330 for the Victor L. Streib Papers, 1908-2012, Undated. M.E. Grenander Department of Special Collections and Archives, University Libraries, University at Albany, State University of New York. I was overjoyed to find the work of Hearn---a tremendous help in the initial dataset construction.

Initially I heavily depended on Dr. Baker's book, Women and Capital Punishment, specifically Chapter 2, which contains a crucial table. This table was, honestly, the very first data I entered into my Excel spreadsheet. I then compared those names to the ones in the Espy file to initiate my project. The Death Penalty Information Center's data was also used. I chose to begin with Velma in 1984. My goal was to double-check with the DPIC to ensure all data matched Baker's dataset. The only significant discrepancy I found, and am still working on, concerns Judy Buenoano. While she is also marked as White, there were some questions about her ethnicity that still need verification. Although current citations suggest she identified as White, I cannot definitively confirm her race until I receive further confirmation. For this reason, her name is also italicized, which signifies a correction to the original data in the "Additional Race" column.

My initial project idea, or rather, my book idea when starting a few years ago, wasn't to highlight discrepancies between the Espy file and the Espy papers. In fact, I had hoped to solely rely on the Espy file. It was only after encountering Professor Streib's, Hearn's, and Baker's research and conducting further investigations myself that I realized how incredibly inadequate and often sexist existing datasets for women's executions were. The Espy Papers were used as the primary starting point for variable completion.

I also relied on other research that has been critical of the Espy file, and this work is cited in my project. The Blackman and McLaughlin (2011) study, for instance, was particularly helpful in identifying names. To the best of my knowledge, the Black and Mclaughlin study has been the only real attempt to point out Espy file inadequacies: The Espy File on American Executions: User Beware. Homicide Studies, 15(3), 209-227. https://doi.org/10.1177/1088767911418054. I found that publishing papers that critique the file were met with almost immediate suspicion from journal editors. Not an entirely new experience for me as I remember the days of explaining why studies about sexual violence among Queer folk "matters." To get back to it, my research simply involves searching for one execution in a primary resource, such as the Library of Congress, and then following the leads. During these searches, I occasionally stumble upon another potential execution, especially in compensation cases (collected also by Baker). These cases require further follow-up and can be challenging to research in some areas. In certain states, for various reasons, compensation case records are not visible online, at least not as readily as they once were.

The variables (for most up-to-date notes, see dataset)

Race and Ethnicity Variables

Variable three, which I may occasionally hide from the sheets for clarity, is "Additional Race." I've marked Judy Buenoano's additional race as "Hispanic?" to remind myself to look this up, though all sources seem to indicate she identified as White. The "Additional Race" variable is used for multiracial individuals or to capture additional racial/ethnic characteristics. This variable is still in progress.

While there was considerable agreement between the Baker and Espy datasets, Baker's other work provides more detailed racial characteristics. This allows for the addition of racial attributes, especially for categories like Native American (see Baker's Registry of Known American Indian Executions 1639-2006), where tribal affiliation is known. Currently, I'm leaving most "Additional Race" fields blank because, for most individuals, they were marked as belonging to a single race. Again, Variable 3 is certainly a work in progress.

Age and Case Particulars

The variable I'm currently most interested in is age, as it's significantly under-collected. I'm finding it challenging to ascertain ages, particularly for individuals executed in earlier years. I anticipate needing to consult additional genealogical sources to determine birth dates and verify ages. For example, June Champion, the first woman executed in the United States, has surprisingly little information about her case particulars, including her age.

Crime Variables and Victim Information

The two variables "Crime Executed as per ESPY file" and "Crime Executed as per Baker" can overlap, sometimes present differences, and sometimes complementary information. For instance, in cases involving both witchcraft and murder, like Kendall Elizabeth's case, further investigation is warranted. In many cases, the Espy file overrepresented "murder" as the crime executed for so using Baker data provides more detail.

One area I'm focusing on, which I appreciate about the Baker file, is trying to ascertain the victim's identity. Given the numerous cases of infanticide in the Espy file (where they were often marked as "murder"), it's notable that the Espy file (of course I'm referring to the data entry personnel) did count "concealing a birth." This is an important historical component of why women were executed during that period. Therefore, combining "concealing a birth" with "infanticide" (as per Baker's coding) provides a more comprehensive understanding of the cases. Women were executed for known stillborn births--a subject covered by more than a few notable researchers.

Additional Sheets

For references and keeping track of any additional executions I might come across in my research. A new addition perhaps? This was taken directly from "The Witchcraft Delusion in Connecticut" provided by the Gutenberg Project:

MARY JOHNSON. Windsor, 1647.

There is no documentary or other evidence to show that Mary Johnson was executed for witchcraft in Windsor in 1647. The charge rests on an entry in Governor Winthrop's Journal, "One ---- of Windsor arraigned and executed at Hartford for a witch." WINTHROP'S History of New England (Savage, 2: 374).

No importance would have attached to this statement, which bears no date and does not give the name or sex of the condemned, had not Dr. Savage in his annotations of the Journal (2: 374) asserted that it was "the first instance of the delusion in New England," and without warrant added, "Perhaps there was sense enough early in the colony to destroy the record."

In all discussions of this matter, it has been assumed or conceded (in the absence of any positive proof), by such eminent critics and scholars as Drake, Fiske, Poole, Hoadley, Stiles, and others, that Winthrop's note was based on rumor or hearsay, or that it related to the later conviction and execution of a woman of the same name, next noted, and the errors as to person, time, and place might easily have been made.

Research informed variable constructions

All alterations submitted to this project are considered suggestions until the book manuscript is complete. This website is open to contributions from any academic or community researcher (for a fuller definition, see Patricia Hill Collins’s Black Feminist Thought (1990)). Formal academic credentials are not required; however, contributors must provide credible evidence and reliable sources for any information they add. Numerous sources already exist for consultation, and this site highlights many of them. If you know of relevant books, archives, library collections, or other materials not yet included, please share them.

I hope eventually to see this project in book form. However, given its scope, it is essential that any issues or inconsistencies be identified early, especially considering the gravity of the subject matter and the potential impact of the data.

At present, I am releasing potential variable constructions gradually, as I am still completing the overarching data framework. I am developing variables that account for experiences such as past trauma, sexual assault, domestic violence, and mental and emotional health. It was important first to establish the backbone of the dataset by writing measurement descriptions for variables originally created in the Espy file and in Baker (2016). Additional variables are not solely of my creation; rather, they adapt or build upon the scholarly work of others to better capture women’s experiences through quantitative data. All variables are fully sourced and referenced, using original materials whenever possible (e.g., court records, correctional documents, autopsies, and other primary sources).

The formatting of the codebook follows conventions used by the ICPSR, as one anticipated outcome is the potential submission of this dataset to ICPSR in addition to its availability on the OSF website. Both the data structure and the dataset description follow ICPSR guidelines whenever possible: https://www.icpsr.umich.edu/web/pages/deposit/index.html.

Page updated

Report abuse