What information is included in our dataset?
Our data set is the Survey of Scottish Witchcraft by Julian Goodare, Lauren Martin, Joyce Miller and Louise Yeoman, published in 2003. The database consists of three levels: the accused, case description, and trial process (a) with references (b). The accused level contains biographical information such as name, sex, residence, and occupation for each accused witch. Goodare et al. found a total of 3,837 people who were accused of witchcraft in Scotland, but 625 of them were unnamed people or groups in the database. The case description level covers many details regarding the accusation: the denunciation, the investigation, and the way the case ended. The case level also goes into imprisonment length and outcome, characterization of cases (i.e. demonic possession), and reasons for accusation. Some witches may have had more than one case, but instead of creating two separate cases, Goodare et al. decided to create two trial processes. The trial process consists of three different types: central trial, local trial, and mixed trial. Some types of trial evidence include mentioning a witch in the trial of someone else, as well as information about the sentence. The reference section lists the primary references that were used and examined by researchers for each individual accused case. The title of the work along with specific volumes and page numbers are also included in this description.
Where was our dataset derived from?
The data was derived from printed works by Larner et al, George F. Black, and Stuart MacDonald. These authors had previously done surveys on Scottish witchcraft trials, which allowed the project members to assess and then compile information into an electronic database. The original sources include both secular and ecclesiastical records. The secular sources being: A Source-Book of Scottish Witchcraft compiled by Christina Larner, Christopher Hyde Lee, and Hugh V. McLachlan in 1977; High Court of Justiciary Books of Adjournal; Circuit Courts books; Privy Council material; Parliamentary material; printed local records or other people’s surveys of local records; Register of Privy Council (RPC), additional manuscript Privy Council minutes, Acts of the Parliament of Scotland (APS) and the Committee of Estates; and ecclesiastical records: all surviving presbytery records covering the years 1563 to 1736 and kirk sessions from high intensity years.
What information, events, or phenomena your dataset can illuminate?
Witchcraft was a secular crime between 1563 and 1736, and almost all trials occurred in this period. Most trials were authorized centrally, and a record of this authorization usually survives, at least after 1608. So, we can obtain a good statistical sense of the pattern of accusations. From this, the dataset illuminates a significant case of gender inequality in Scottish society in the 16th and 17th century from the accused list with 70-80% being women. In literature about other European countries in this era, this was also a common trend as the sentiments about witch hunts followed the witch hunters’ guide, “Malleus Maleficarum” written by Heinrich Kramer and Jacob Sprenger, Catholic clergymen from Germany. Common social factors such as religion and a judicial system ruled by the church would also reinforce the image of Scottish society and gender injustice which our data seems to paint. Other than that, the various characteristics/case attributes the accused witches are associated with include being involved with fairies, implication by other witches, health consultation, midwifery, unorthodox religious practice and more. These case attributes differ between men and women and could potentially reveal what was deemed as wrong in Scottish society. We feel that the disparities between gender could possibly illuminate medieval Scotland gender stereotypes. Also, the peak period of witch trials was in 1660, the start of the “Great Scottish Witch Hunt” time period where torture was permitted to be used as a part of the legal process. Other than that, the data reveals the methods of torture and types of evidence used in the trials.
What information is left out of the spreadsheet?
The project members identified 3,837 people who were accused of witchcraft in Scotland from 1563-1736. However, since the data was compiled from multiple sources and not every detail was recorded for every accused person, there are blanks in the data. There are 625 records for unnamed people or groups, so only 3,212 of the accused are named. Not all cases recorded whether the witches were executed, trial outcomes, accuser information, and much more. Also, historically most witches were tried in local criminal courts, but the data isn’t able to fully be represented since records from these courts rarely survive. We also struggled with drawing conclusions from verdict and torture data as there were very few records relative to accusation records. The biggest challenge with our data is that there are separate unique ID’s for each level of the trial which complicates the data analysis.
What can’t our dataset reveal?
This data doesn’t provide a complete picture of cultural nuances or shared mentalites that could have impacted these outcomes, for example the “fairies” and how many people in early Scotland associated mental disorders with the supernatural. We can’t argue that these accusations were solely on people who were mentally ill, but it’s something that can be personally noted as we move through the data. The why is often missing in data, and even skewed by those who have the pen/power. We feel that our dataset oncology is limited in giving causal explanations on any changes in the number of cases across different geographical locations or time periods. For instance, external studies showed that economic distress possibly played a significant role in the rise of cases but this cannot be observed from the database. Hence, to aid in understanding the narrative surrounding our dataset it is important to include other sources of research found in peer-reviewed academic papers and articles. In doing so, this not only helps us flesh out the “why” of our data, but also familiarizes us with frameworks that surround witchcraft and how Scotland may or may not prescribe to these similar themes. Our dataset is also missing qualitative data such as the stories of the accused witches, who they were, where they came from or whether they were actually innocent in reality.