CWGK Annotates 1,000 Documents

CWGK editorial staff recently annotated their 1,000th document. They have identified and written short biographies of each person who appears in 1,000 of the more than 10,000 documents that make up CWGK. Although their work continues, they are sharing their thoughts at this milestone in three blogs published by KHS. Read excerpts below, and follow the entire series through the links.

Natalie Smith, “Uncovering Untold Stories of Civil War-era Kentuckians
Editing documents in the Civil War Governors of Kentucky Digital Documentary Edition (CWGK) often means I’m diving headfirst into the grittier aspects of the Civil War. Crime, poverty, starvation and guerrilla attacks only scratch the surface of what Kentuckians endured during this chaotic period in our country’s history.

Emily Moses, “Staff Member Gains Insight from CWGK
Every day that I work on The Civil War Governors of Kentucky Digital Documentary Edition (CWGK) I am in awe of how this digital collection provides insight into the lives of 19th century Kentuckians. I am not a Kentuckian, nor by training am I a Civil War historian; I am a southern historian.

Patrick Lewis, “Annotation Requires Writing for People and Machines
Our biographies work. They build upon one another to create hundreds of thousands of discrete records of historical events large and small. Each treated equally. The world we capture in our biographies can be set into motion; viewed from the perspective of a town, of a day, of people who journey together on a specific steamboat. The number of these stories that we encode, both mundane and the world-changing, are almost limitless. The scale of this data is so great that we can’t yet fully imagine how we are going to use it. Has the historian who will build the system that starts and stops this network in time or peeks into the totality of a community on a critical month or day been born yet?

CWGK in The Federalist

What expectations did people have of local, state, and federal governments? Who were the faces of governance in their communities? How did they conceive of justice and equity? How did they understand the interaction of branches and levels of government, and how did they play governing institutions off of one another to secure the outcomes they desired?

In the Fall 2018 issue of The Federalist, the newsletter of the Society for History in the Federal Government, CWGK Project Director Patrick Lewis reflected on the important and relevant questions that the project has raised — both in the materials that it has found, published, and annotated and also in the process of managing a program within state government.

From using social networking to discover local power brokers operating outside the formal channels of power to appreciating the inability of antebellum institutions to cope with the overwhelming crisis that secession and Civil War brought to Kentucky society, CWGK provides a new research path forward for historians. How did people understand their government before the war and, when the conflict came to their doorstep, what expectations did they have for government intervention and assistance?

I have developed a profound empathy for both the plaintive citizens bringing horrifying tales of death, crime, sexual violence, destitution, and starvation as well as for the representatives of government at all levels who are chronically unable to muster sufficient resources to address the systemic problems they saw. It is easy to see the
Civil War as a crisis of elected government—at a legislative,
gubernatorial, Congressional, and especially Presidential level—but I have come to appreciate the war as it drug down an underprepared and underpowered civil service under the weight of modern, total war. The antebellum systems buckled underneath the crisis. That book is far more complicated to write than a conventional political history and far less marketable than a new battle history. That book about the slow collapse of governmental systems under unforeseen external stress might also b far more relevant to a moment when the national coffers have been drained by years of military conflict and faith in the capacity of electoral politics to address the day-to-day issues facing the citizenry is critically low.

Access a PDF of the full article here, or read the full issue at The Federalist.

Using CWGK Annotations

The Civil War Governors of Kentucky (CWGK) team is happy to announce the launch of the new, annotated CWGK website! The updated and expanded site publishes for the first time 350 fully annotated documents and combines the previously launched Omeka platform with the power and versatility of our annotation tool—Mashbill—to produce complex social networking visualizations for each entity (person, organization, place, and geographical feature). The CWGK team worked with Brumfield Labs and Dazhi Jiao to complete this latest digital publishing platform.

Let’s explore the entity page of Governor James Fisher Robinson. You can see Robinson’s social network and a visualization of this network—for example, his connection to G. F. Cook (circled in black).

And looking at the entity page for G. F. Cook, we see his connection to Governor Robinson.

The legend on the left depicts the different types of entities you will see in the visualization and the different types of relationships that link them together.

Back to Robinson’s entity page, there are many more important pieces of information. First, you can see the full biographical entry for Robinson and the citation for the sources consulted in writing his biography.

Below the visualization are a series of tabs that will give you access to additional information and tools about each entity. There are four tabs: Metadata, Citation, Documents, and Download.

The metadata tab will give you access to the entities birth date, death date, gender, race, and entity type.

The citation tab will give the full entity citation for the convenience of the researcher.

The documents tab will give you a list of EVERY document that this particular entity is linked to throughout the CWGK website. The list is quite long for Robinson.

The download tab will allow you to download the XML code for that particular entity.

This is just the beginning of publishing fully annotated documents and visualizations on the CWGK website. Eventually, tens of thousands more documents and hundreds of thousands more entities will be published. Updates to the site will appear continually as the editing process continues and each document is completed.

So stay tuned and visit often!

CWGK Annotation is Live!

2017 wraps up with a triumph for CWGK. The project published its first 350 annotated documents, with each of the people, organizations, places, and geographical features mentioned in the documents linked in the text. The basics of CWGK’s document viewer interface look the same, with the addition of the links to the annotated entities highlighted as they appear in the transcription.

When a user clicks on an highlighted name, they will navigate to the entity’s page. Whatever CWGK researchers have found about the individual will be written in a short biographical statement, with a full citation just below it. Most strikingly, for people and organizations, the user will see a social network visualization drawn from the relationships found in CWGK documents. These can range from the simple:

to the spectacular:

These annotated documents aren’t the only new edition to the CWGK site. The project has also updated over 1,600 transcriptions, delivering users the most reliable representation of the text possible.

Stay tuned for more, too. CWGK’s new system allows for rolling publication as new transcriptions and annotations are approved. From this time forward, CWGK will be an ever expanding and interconnecting network of primary and secondary material that allows new and deeper access to the lives of everyday people caught in the middle of the country’s greatest conflict.

A great deal of this editorial work was funded by a grant from the Scholarly Editions and Translations program of the National Endowment for the Humanities. Development of the annotation and visualization systems were funded by two grants from the Publishing Historical Records in Documentary Editions program of the National Historical Publications and Records Commission.

CWGK Annotation Preview at DH2017

CWGK’s development partners at Brumfield Labs have been working on a series of NHPRC grants to develop MashBill, CWGK’s annotation management system. Through the work of NHPRC-funded Graduate Research Associates, CWGK has annotated approximately 1,200 documents to date, identifying over 8,000 unique people, places, organizations, and geographical features.

At the Digital Humanities 2017 conference in Montreal, Brumfield Labs presented a paper co-authored by CWGK staff, “Beyond Coocurrence: Network Visualization in the Civil War Governors of Kentucky Digital Documentary Edition.”

Read a full recap of the presentation here, complete with fascinating visualizations of CWGK annotation data drawn from MashBill. The recap was named an Editors’ Choice story by Digital Humanities Now in August 2017.

Graduate Research Associates 2016-17

Overview

The Kentucky Historical Society seeks eight Graduate Research Associates (GRAs) familiar with 19th century United States history to write short informational entries for the Civil War Governors of Kentucky Digital Documentary Edition (CWG-K). GRAs will receive a stipend of $5,000 each and can work remotely from their home institutions.

Each GRA will annotate 150 assigned documents each. Each GRA must be a graduate student in at least the second year of a M.A. program in history or a related humanities discipline. In accordance with its commitment to facilitating relationships between history practitioners and organizations in Kentucky and nationally, KHS hopes that these GRA positions will help advance the professional skills of early-career historians in Kentucky and elsewhere. Preference will be given to candidates who are enrolled in graduate programs in history at Kentucky universities, though applicants worldwide are encouraged to apply. These positions are funded by a grant from the National Historical Publications and Records Commission (NHPRC), a branch of the National Archives.

CWG-K is an annotated, searchable, and freely-accessible online edition of documents associated with the chief executives of the commonwealth, 1860-1865. Yet CWG-K is not solely about the five governors; it is about reconstructing the lost lives and voices of tens of thousands of Kentuckians who interacted with the office of the governor during the war years. CWG-K will identify, research, and link together every person, place, and organization found in its documents. This web of hundreds of thousands of networked nodes will dramatically expand the number of actors in Kentucky and U.S. history, show scholars new patterns and hidden relationships, and recognize the humanity and agency of historically marginalized people. To see the project’s work to date, visit discovery.civilwargovernors.org.

Scope of Work

Each GRA will be responsible for researching and writing short entries on named persons, places, organizations, and geographical features in 150 documents. Each document contains an average of fifteen such entities. This work will be completed and submitted to CWG-K for fact-checking before June 30, 2017.

Research and writing will proceed according to project guidelines concerning research sources and methods, editorial information desired, and adherence house style. This will ensure 1) that due diligence is done to the research of each entity and 2) that information is recorded for each item in uniform ways which are easy to encode and search.

All research for the entries must be based in primary or credible secondary sources, and each GRA is expected to keep a virtual research file with notes and digital images of documents related to each entry. These will be turned over to CWG-K at the completion of the work. CWG-K will fact-check all entries for research quality and adherence to house style. CWG-K projects an average rate of one document annotated per two hours of work. Each GRA may expect to devote approximately 300 hours to the research—though the actual investment of time may vary.

Each GRA will work remotely. Interaction with the documents and the writing of annotations will take place in a web-based annotation tool developed for CWG-K, which can be dialed into from any location. CWG-K will make use of online research databases to make its work efficient and uniform. Other archival sources may be of value but are not required by the research guidelines. Securing access to the paid databases required by CWG-K (Ancestry.com, Fold3.com, and ProQuest Historical Newspapers: Louisville Courier Journal) is the responsibility of the GRA. If regular institutional access to these databases is not available to the GRA through a university or library, it is the responsibility of the GRA to purchase and use a subscription to these databases. KHS will not reimburse the GRA for any travel, copying, or other expenses incurred in CWG-K research.

In order to maintain quality and consistency as well as to foster a collegial and collaborative work culture, CWG-K will conduct weekly virtual “office hours” via Google Hangouts, during which GRAs are required to dial in, ask questions of staff, share expertise and research methods, and make connections with their peers at other universities. Virtual attendance at these office hours is mandatory, and multiple sessions may be offered to accommodate schedules.

The Kentucky Historical Society will hold copyright for all annotation research as work for hire.

Evaluation Criteria

A proposal should consist of at least a narrative statement of professional ability in the form of a cover letter, a CV, and two letters of recommendation. Additional supplementary materials that demonstrate capacity in the evaluation factors may also be included. Applications are due by September 16, 2016 to Tony Curtis, tony.curtis@ky.gov.

The Kentucky Historical Society will evaluate the proposals based on the following factors:

Research Experience (70 points): Describe your familiarity with research in 19th century U.S. history. Describe some projects you have undertaken. What sources have you used? Have you been published? Have you interpreted historical research in forms other than a scholarly peer-reviewed publication? How does the proposed research project differ from those you have undertaken in the past? Describe your familiarity with the strengths and weaknesses of online research databases such as Ancestry.com, Fold3.com, ProQuest, and Google Books.

Project Experience (30 points): Describe any work you have done in the editing of historical documents. Discuss how a project such as CWG-K maintains balance between thorough research and production schedules. Have you worked on other collaborative projects in the field of history or otherwise? Describe your ability to meet deadlines and regulate workflow. Describe your understanding of and/or experience with the Digital Humanities. From what you know of the CWG-K project, how does it fit with current trends in the field? What do you hope to gain from working on the CWG-K project?

Institutional Affiliation (10 points): Additional points are available to applicants who are enrolled in graduate programs at Kentucky universities. Applicants claiming this status should discuss how they will use this experience to help build and sustain relationships among history organizations across the state and articulate why such relationships are valuable. This does not imply any relationship between KHS and the educational institution.

Making Connections to the Past

By Stefanie King

Civil War Governors‘ new Early Access web interface provides researchers access to thousands of transcribed Civil War-era documents that bring new voices into conversation with historians. But how does the project move beyond the texts to provide researchers new levels of access to the lives of 1860s Kentuckians? Early Access is an important achievement for the project, more than five years in the making. Yet the next phase of work will push the boundaries of digital scholarship by using these documents to map the social network of Civil War-era Kentucky.

This summer, Civil War Governors is trying to understand how dense and interconnected a network the documents will allow us to construct. So we chose 21 documents to be our laboratory experiment before moving on to the 23,000 identified so far.

Caroline Chron Connections

Visualizing Caroline’s Social World

Those 21 documents contained 440 identifiable entities (IDs), including people, places, and organizations. Some of the people mentioned in the documents are easy to identify, such as the governors or other prominent members of society. Others are harder. Caroline Dennant, from The Caroline Chronicles, is is difficult to research due to the biases of the historical record, stemming from her life as an enslaved woman and then as a contraband. Most of what we know about Caroline comes from the context of the documents in which she is mentioned.

Other people are difficult to identify because of how they appear in a document—for example, the “german woman” referred to in Caroline’s court case will be tough to identify because the information about her is vague (KYR-0001-004-0131). Similarly, many of the people mentioned in the documents are difficult to identify because their full name is not included. A Kenton County petition signed by “R Mann” is a start. But is he Robert Mann or Richard Mann, both of whom lived in the county at the time? (KYR-0001-020-1405). As frustrating as it can be, though, the process of identifying little-known historical actors includes some interesting discoveries as well, such as Willis Levy’s neighbor and brother, James, who was a lightning-rod maker.

Understanding how the people are connected is a challenging task as well.  One obstacle, again, stems from the limitations of the historical record. For example, we know Rev. John L. McKee met with Caroline Dennant on multiple occasions, primarily providing her with religious counsel. But were they close friends, or merely acquaintances? Without further information from the people themselves, we can only determine the nature of that relationship as revealed in the extant documents. Furthermore, the documents do not tell us why Rev. McKee decided to help Caroline. Did Caroline seek out his counsel? Did a member of Rev. McKee’s congregation request that he become involved in Caroline’s case? Did McKee and Caroline already know each other somehow? Although we know there was a relationship between Rev. McKee and Caroline Dennant, we do not know how the relationship began, or what became of that relationship.

This leads to the second step in understanding the social network in Civil War-era Kentucky, which is categorizing types of relationships. Some relationships are easy to understand: Willis and Anne Levy were married; Blanche Levy was the child of Willis and Anne Levy; Anne Levy and Josephine Lynch were sisters.

Other relationships are not as easy to categorize. For example, numerous concerned citizens petition Governor Bramlette, asking that he pardon Caroline Dennant. But what is the nature of the relationship between these petitioners and Caroline? Are the petitioners friends of Caroline? Acquaintances? Some of the petitioners, such as John G. Barret, did not even know Caroline, and Caroline may not have known all of the people who petitioned the governor on her behalf (KYR-0001-004-0129). To complicate things further, nine of the jurors involved in Caroline’s case petitioned Governor Bramlette to pardon Caroline. The jurors who convicted Caroline of murder, then petitioned that she be pardoned. What type of relationship does that indicate? “Juror” or “petitioner” may not constitute a relationship, but the action of serving on the jury or signing a petition does establish a connection between them.

Deciding how to classify a huge range of human relationships into a handful of regularized relationship types is a tricky process that balances usability and nuance, generality and specificity. If Civil War Governors gets it right, researchers will be able to discover new patterns that would otherwise not be apparent, as well as have access to a new biographical encyclopedia of everyday people of all walks of life in Kentucky history. The project will allow researchers to visualize Civil War-era Kentucky by revealing the connections that underpinned this nineteenth-century world.

Stefanie King is a Ph.D. student at the University of Kentucky and a summer 2016 intern at the Civil War Governors of Kentucky Digital Documentary Edition.

NHPRC Support

The Civil War Governors of Kentucky Digital Documentary Edition and the Kentucky Historical Society are proud to announce that Civil War Governors was awarded $62,400 in the May 2016 cycle of National Historical Publications & Records Commission (NHPRC) funding.

This is the second grant Civil War Governors has recieved from NHPRC. The first grant supported the digital publication of Early Access, an Omeka interface that will host 10,000 Civil War-era documents. This site will allow users to access 10,000 digital document images and transcriptions, and sample the rich content that Civil War Governors will deliver. Civil War Governors true impact on scholarship, however, will be in annotation. To the extent possible given the restrictions and biases of the historical record, Civil War Governors will identify, research, and link together every person, place, and organization found in its documents. This web of tens—perhaps hundreds—of thousands of networked nodes will dramatically expand the number of actors in Kentucky and U.S. history, show scholars new patterns and hidden relationships, and recognize the humanity and agency of historically marginalized people. The network of identified and annotated people, places, businesses, government agencies, and military units, will come as close as possible to a historical reconstruction of mid-nineteenth century society as it was lived and experienced in wartime Kentucky.

This second grant will support that next phase of work: publishing an annotation interface of 1,500 fully edited and linked documents. This important step to the full Digital Documentary Edition of a projected 40,000 documents will help reconstruct the lost lives and voices of tens of thousands of Kentuckians who interacted with the office of the governor during the war years. The diverse and largely unknown lives of the people captured in the Civil War Governors documents—generals and politicians, prostitutes and plantation mistresses, free African American professionals and “contraband” refugees—as well as the social, economic, political, and geographic networks they allow us to visualize and understand, are what Civil War Governors hopes to capture and deliver to the public.

This will be an important test of the Civil War Governors vision of a digital research environment within which a user can encounter the past multi-dimensionally through the documents and the powerful annotation network that links the documents together. In this document-driven historical ecosystem, users can explore intuitively—moving seamlessly through seemingly disparate historical themes, events, and topics; breaking into the plane of social and geographic space to understand the deep patterns that underlay the issues raised in a text or set of texts. Through it, Civil War Governors will understand how project staff must balance all phases of editing work as well as how a variety of users will navigate the research platform. Early Access shows the public a tantalizing sample of Civil War Governors content, but the next phase will demonstrate how Civil War Governors will shape the ways researchers, students, and teachers will explore the past in the future.

The Caroline Chronicles: A Story of Race, Urban Slavery, and Infanticide in the Border South – PART III

The Caroline Chronicles:
A Story of Race, Urban Slavery, and Infanticide in the Border South

“Part III – The Defense’s Case”

By Patrick A. Lewis

For those of you who missed previous installments, we’ll begin with a very brief rundown of Caroline’s story to this point. (A full accounting of the events that led to her trial for infanticide is still available here.) In 1862 Caroline Dennant, a Tennessee slave, was brought to Louisville, Kentucky, as war contraband by Don Carlos Buell’s army—she was subsequently arrested as a fugitive slave and placed in the home of Willis and Anne Levy—a few months later, Blanch, the Levy’s toddler-aged daughter died of strychnine poisoning—Caroline was soon after charged with murder, convicted, and sentenced to death. This and last week’s installments are written from the perspective of the prosecution and the defense in the matter of Caroline’s petition for executive clemency (and may or may not reflect our actual positions on her case!).

As the prosecution alleges, there is little the defense can do to refute the circumstantial evidence against Caroline. She had been held to labor as a servant and nurse in the home of the Levys. Willis Levy did acquire, distribute, and store a large amount of strychnine. After the child’s death, Caroline was seen to have facial expressions and otherwise behave in ways to which sinister motives were later assigned by witnesses. While the defense concedes this circumstantial evidence, it entirely rejects the fanciful and conspiratorial theory of the (so-called) crime advanced by the prosecution.

Yet to secure the conviction in the trial at the May 1863 term of the Jefferson Circuit Court, the defense knowingly suppressed the extent to which Willis Levy “spread enough strychnine (or poison) to kill a regiment of men” in and about his premises. Evidence freely offered by the neighbors and family of the Levy family since the time of the trial now begs reconsideration of the case. The defense appeals to the clemency of the executive for a pardon on the following grounds:

One. That having resided in Louisville less than six months before the death of the child Blanch Levy, “in a strange place without any one to advise with” except defense counsel hastily assigned her case and without adequate time to prepare, Caroline was unable to secure witnesses for her defense at the trial.

Two. That the witnesses for the prosecution, namely Anne and Willis Levy, did not testify to the full extent to which Willis Levy spread strychnine about his premises. Only two occasions were established in evidence by Willis Levy, and Caroline could swear to no more. “Your petitioner will now state one important fact which was not developed on the trial, Mr Levy put out the poison on more than two occasions; he put it out many times to kill Dogs & Cats, & it was never taken up, & what became of it no one knows.”

Three. That the testimony of Raymond and Josephine Lynch—neighbors and in-laws to the Levys, uncle and aunt of the deceased Blanch Levy—establishes the true extent of Willis Levy’s indiscriminate and dangerous application of strychnine in and around his and his neighbors’ property. Josephine Lynch swears that “Mr Levy put out the poison every night for a great while I would think a hundred times” over a span of time “from fall to spring.” Moreover, Mrs. Lynch herself had been “very uneasy many time for fear that my children would get some of the poison I alwaise thought Mr Levy was very reckless about throwing out poison.”

Four. That the prosecution argues against accidental ingestion of the poison in the yard from the fact that no pieces of poisoned meat were found in the stomach of the deceased Blanch Levy.

Five. That testimony developed on the trial and that subsequently sworn to by Josephine Lynch establishes that a considerable amount of strychnine was spread in the yard and neighbors’ yards by means other than on meat, including but not limited to on grains designed to kill birds and loosely distributed in and around the privy.

Six. That Mrs. Levy grasped the extent to which her husband had indiscriminately spread poison in and around the Levy house. Immediately after the child’s death Mrs. Levy threw out a “bucket full of parched coffee that was bought from the soldiers,” believing it to be tainted with the poison.

Seven. That if Anne Levy was made sick by coffee on the morning the child died, this was from Willis Levy unwittingly contaminating the household coffee supply with strychnine as part of his campaign to eradicate vermin.

Eight. That if the true extent to which Willis Levy indiscriminately scattered strychnine in and around his own property and that of his neighbors had been known at the time of the trial, Caroline’s conviction would not have been sought by the prosecuting attorney. Louisville City Attorney William G. Reasor attests that “from strong circumstances made known to me since that trial, I feel that Executive clemency will have been worthily bestowed if she be fully pardoned.”

LevyNine. That if the true extent to which Willis Levy indiscriminately and dangerously scattered strychnine in diverse methods and in diverse locations in and around his own property and that of his neighbors had been known at the time of the trial, Caroline’s conviction would not have been secured by the jury. Nine of the gentlemen of the jury who tried her case—L. A. Civill, W. O. Gardner, John Sait, Joseph Griffith, Thomas Schorch, Samuel Ingrem, R. H. Snyder, William K. Allan, and E. P. Neale—have signed a sworn statement asking to overturn the verdict and sentence they rendered.

All this the defense presents as evidence for Caroline’s innocence in the death of the child Blanch Levy. The defense will not—as it believes it has grounds to do—pursue the argument that Caroline’s service in the Levy household was in violation of the Confiscation Act of July 17, 1862, which provides that “all slaves of persons who shall hereafter be engaged in rebellion against the government of the United States “shall be forever free of their servitude, and not again held as slaves” and that “no slave escaping into any State, Territory, or the District of Columbia, from any other State, shall be delivered up, or in any way impeded or hindered of his liberty” regardless of the laws pertaining to enslaved persons and persons of African descent in that state, territory, or district.

The defense reiterates that given the circumstances of the defendant and her insecure position in Louisville, the evidence presented in this petition was unavailable to Caroline and her counsel at the time of the trial.

If all that were introduced in this petition were this new testimony, the defense would feel confident in their expectation of His Excellency’s clemency, but having in hand the sworn statements of the prosecuting attorney and the jury, the defense feels that the pardoning power would be justly used in the case of Caroline. The premises considered, the defense asks that His Excellency Governor Bramlette issue a full and unconditional pardon.

The Caroline Chronicles: A Story of Race, Urban Slavery, and Infanticide in the Border South – PART II

The Caroline Chronicles:
A Story of Race, Urban Slavery, and Infanticide in the Border South

“Part II – The Prosecution’s Case”

By Matthew C. Hulbert

For those of you who missed last week’s installment, we’ll begin with a very brief rundown of Caroline’s story to this point. (A full accounting of the events that led to her trial for infanticide is still available here.) In 1862 Caroline Dennant, a Tennessee slave, was brought to Louisville, Kentucky, as war contraband by Don Carlos Buell’s army—she was subsequently arrested as a fugitive slave and placed in the home of Willis and Annie Levy—a few months later, Blanch, the Levy’s toddler-aged daughter died of strychnine poisoning—Caroline was soon after charged with murder, convicted, and sentenced to death. This week’s installment—and next week’s—are written from the perspective of the prosecution and the defense in the matter of Caroline’s petition for executive clemency (and may or may not reflect our actual positions on her case!).

The charge against Caroline revolves around a web of evidence, the majority of which is deemed circumstantial. On the surface, this would appear to weaken the state’s case. However, in instances where such a preponderance of circumstantial evidence points to the guilt of an individual, such as in this instance, logic will not allow us to be swayed by the unreasonable possibility of coincidence. When Caroline’s case is dissected, thread by thread, you will see that she not only committed an act of premeditated murder against a defenseless and innocent child to punish her temporary guardians—but that she potentially did so as part of a broader, though admittedly poorly-conceived, plan to escape from the Levy’s care and to circumvent the possibility of a return to bondage in Tennessee.

Here are the main pillars of the state’s case, laid out as individual items:

One. We know based on the autopsy performed by Dr. Jenkins (a professional chemist) that Blanch Levy died as the result of strychnine poisoning, with significantly more than a fatal dose of the substance found in her stomach. Both the location (stomach) and quantity of the person underscore that the substance was ingested directly and not absorbed through skin contact, accidental or otherwise.

frog stomachTwo. We know based on her own petition for executive clemency that Caroline knew the whereabouts of the strychnine kept in the Levy household and that through the testimony of Annie Levy—that the trunk containing the poison was not locked—that Caroline had ready access to the substance whenever she pleased. The defense does not dispute either of these points.

Third. Caroline had double-motive for killing Blanch Levy: revenge and personal gain. On one hand, Willis Levy became increasingly critical of Caroline’s poor behavior. The record indicates that through negligence, Caroline was responsible for damaged fruit trees and for the fouling of a newly-washed fence. On at least one occasion, the defendant reports that Willis Levy noted that he would like to whip Caroline—but the defendant did not testify to any instances of physical abuse taking place in the Levy household. Moreover, so long as she remained under the Levy’s roof, Caroline ran the risk of being returned to permanent bondage in Tennessee. As she had been declared a fugitive slave and arrested, the Levy’s were essentially providing her with a temporary home until her former master claimed her or until she could be sold at auction by local authorities.

Four. Annie Levy testified that on the day preceding the death of her daughter, she arrived home to find that the trunk containing the poison had been clearly disturbed. Caroline denied having opened the trunk, but did not deny that the trunk itself had been moved and its contents shifted.

Five. We know that in conjunction with the trunk having been disturbed, Annie Levy mysteriously fell ill with very mild symptoms indicative of strychnine poisoning—no doubt after consuming a dinner prepared by Caroline—and was still ill the next morning when she and the victim arrived late for breakfast. Caroline’s testimony does not dispute that for the first time in her entire tenure with the Levy family, she prepared and poured Annie Levy’s morning coffee. The defense does not dispute that Annie Levy noted that the coffee had an off taste and she did not finish it.

Six. We know from multiple lines of testimony that the victim, Blanch Levy, was in the sole care of Caroline in the moments preceding her death and that, for the time before she was given into Caroline’s sole care, she exhibited no signs of illness or poisoning consistent with the consumption of strychnine.

Seven. According to the testimony of Annie Levy, when Caroline entered her bedroom to state that Blanch was acting strangely (read: convulsing and choking to death in the front yard), the defendant did so slowly, without any hints of emotional distress or surprise at the events then unfolding. In connection to this lack of emotional distress, on more than one occasion, witnesses saw Caroline look at the child’s corpse and smile.

Eight. Immediately following Blanch’s death, witnesses report that, in the evening, Caroline walked to the gate of the Levy’s front yard and looked around. She had not previously been known to visit the gate in the evenings. The importance of this point will be brought to light later in the prosecution’s case.

Nine. When Caroline realized that Blanch had not been immediately interred, she became increasingly anxious concerning whether or not an autopsy would be performed, reportedly even asking Annie Levy several times when, precisely, the girl’s body would be buried.

Ten. Court documents show—and the prosecution concedes—that Willis Levy did, shortly before his departure on a freight trip, distribute small pieces of beef tainted with strychnine poison to kill local dogs and birds. However, as is also noted, Levy put this poisoned bait under the homes of his neighbors—while Caroline’s petition for clemency highlights that Blanch died just three feet from the kitchen door of the Levy’s home.

With these statements in mind, the prosecution’s theory of the crime is as follows:

While living in a constant state of paranoia—fueled by her fugitive status—Caroline quickly grew tired of working for Willis Levy and for waiting for her former master to materialize at any moment with the intention of dragging her back to bondage in Tennessee. As such, with knowledge of how to use strychnine poison and knowledge of its location in the Levy household, Caroline waited until Willis Levy had left for extended business trip and first targeted Annie Levy. Annie’s dose wasn’t fatal—though it might have been had she finished her coffee—but it was enough to induce sickness. With the child’s mother sick in bed, Caroline had sole control of Blanch. The timing of Willis Levy’s absence, the disturbance of the trunk, Annie’s sickness, the coffee incident, and Blanch’s demise in Caroline’s custody are simply too damning to write off as a coincidence. With no other adult witnesses present, Caroline fed the toddler significantly more than a fatal dose of strychnine. Following Blanch’s death, with didn’t seem to phase Caroline emotionally, she behaved with increasing strangeness; first, concerning the autopsy and burial and the child; and, second, checking the Levy’s gate in the evenings.

The defense will likely raise two primary points of defense on Caroline’s behalf. One: that she was abused and mistreated by the Levy family and killed to protect herself. However, it is well-known that the Levy family actually allowed Caroline’s husband, a contraband slave who lived with their in-laws, to spend the night with Caroline and that she, herself, did not testify to any abuse mistreatment from Levy other than harsh words. Two: that Blanch was poisoned through the negligence of her father, known in the neighborhood for poisoning animals, and that Caroline, as a homeless, African American slave, and as a defenseless woman, became Levy’s scapegoat. The logistics of the case, however, mainly the quantity of poison found in Blanch’s stomach (and the absence of the beef cubes used by Willis Levy) and the physical location of her death discounts this possibility. Furthermore, the sheer quantity of poison found in Blanch’s stomach by the attending physician means that Caroline would’ve had to watch the child ingest multiple pieces of poisoned animal bait and done nothing.

Much more likely is that Caroline waited until Willis Levy—who was more observant of her misbehavior and thus much harder to poison—had left home for an extended period of time. She then attempted poison Annie, who would presumably have died in her sleep that first evening. When that didn’t work, she again tried to poison Annie and also successfully poisoned Blanch. Caroline then checked the gate each evening because, in all probability, she was waiting for her husband to join her in an attempt to flee to permanent freedom. He never came and she was eventually found guilty following a trial in complete compliance with state and local procedures.

In closing, the state is aware that Caroline has doggedly refused to admit guilt and that a number of local citizens—including attorney’s and members of the jury—have joined her plea for executive clemency despite sentencing her to death immediately following the trial. All the state will say concerning this sudden wave of support is that Caroline’s “new friends” are more likely to be using her as a tool to advance their own political causes than to advance the cause of justice. Otherwise, where were they with aid and assistance before she was found guilty and sentenced to hang?

Matthew C. Hulbert is an Assistant Editor of the Civil War Governors of Kentucky Digital Documentary Edition.