Blog 4: Quantitative Analysis and Violence in Historical Materials: The 1641 Depositions
By Dr. Madina Kurmangaliyeva
A new wave of data from the past is helping social scientists uncover the stories of those who had been forgotten. In the VOICES project, we are studying, amongst other things, violence experienced by women in Ireland in the aftermath of the 1641 Rebellion.
In particular, we will compare the violence experienced by women to that experienced by men, and we are doing this across various regions of Ireland, using the 1641 depositions.
The data in the depositions was originally collected by commissioners appointed by the English authorities to document material losses and violence experienced by the Protestant population. Most of the deponents testified because they suffered material losses, but many also testified about violence they had suffered or heard about. The texts of the depositions have been digitised and transcribed by the 1641 Depositions Project and now are available online.
In VOICES, we are transforming a cross section of texts from the depositions into statistical objects in order to provide a comprehensive view on violence, particularly against women.
One challenge in comparing data from depositions across Ireland is accounting for the circumstances of how the depositions were collected in the first place. When we look at the testimonies of people who lived in different places at the time of the rebellion, we want to compare like with like. In other words, we would like the data to be representative of the experiences of mid-seventeenth-century Ireland.
Unfortunately, however, this is not necessarily always the case: where and when the depositions were collected, seems to have had an impact on who was able to testify.
Depositions were not all collected at the same time, nor in the same place, and not even for the same purpose. There are three categories of depositions: (1) Dublin Originals, collected between 1642 and 1647; (2) Bisse (named after Philip Bisse who collected depositions in Munster), collected between 1642 and 1643; (3) Commonwealth, which were collected around ten years after the events occurred. Importantly, each category has a different geographical coverage. For example, Dublin Originals included some deponents who lived in counties that were far away from Dublin (e.g. Cork or Armagh), but most of the deponents came from County Dublin itself, or from neighbouring counties, as shown in Figure 1. The Bisse depositions, however, were collected in Munster, and, hence, most of the deponents came from the southern counties, particularly from Cork.
How far the deponents had to travel, not only changes the number of depositions, but also the nature of the depositions.
To illustrate the change in the nature of the depositions, we broke down the share of depositions that related to violent events such as arson, assault, reports of captivity, killings or deaths (caused by the rebellion), rape, or stripping.
For example, people who travelled further to depose reported violent acts at a higher rate. In the Dublin Originals, for example, 35% of deponents from County Dublin reported at least one act of violence, while this share is 78% for those who travelled from County Cork to depose in Dublin and 89% for those from County Armagh, where the violence was particularly intense. In other words, deponents who lived further away from but deposed in Dublin or Cork were more likely to report violent events. In general, Figure 2 (left panel) shows a positive correlation between the distance the deponents had to travel, and the share of violence-related depositions they provided.
On the one hand, it could be that the violence was particularly bad in counties further away from Dublin. On the other, the positive correlation might be due to selection mechanisms that were at play: the deponents from further away places were most likely refugees fleeing violence, while deponents from nearer Dublin included people who simply wanted to document property and material losses.
Strikingly, the relationship between the share of deponents reporting violence and the distance to County Cork in the Bisse depositions has a similar slope to the one for Dublin Original (Figure 2, right panel). Notable comparison can be made when the same county appeared in both the Dublin Originals and in the Bisse depositions. For example, only 30% of deponents from County Cork reported violence in the Bisse depositions (when the depositions were recorded in Cork) versus 78% in the Dublin Originals. Kings County, however, is much closer to Dublin than to Cork, and hence the share of deponents reporting violence is lower in the Dublin Originals (64%) versus the Bisse (86%) depositions. Kilkenny is somewhat equally distant from both Dublin and Cork, and the share of deponents reporting violence appears to be close: 84% (Dublin Original) and 75% (Bisse). All of this points to the travel distance being important for defining what kind of testimonies were collected but of course other factors such as the presence of enemy troops, the time of year, and the weather may also have had a role to play.
What does this mean for historical research?
First, that we cannot simply mix the depositions from different categories for statistical analysis. For example, we cannot simply compare some key statistics between two counties that are far away from each other as they are less likely to be covered by the same source.
Rather, we have to keep the analysis separate within the same category of depositions, and any comparison across different localities has to account for the travel distances and, whenever possible, also for other circumstances, such as the presence of hostile troops or the time of year. The further we are from Dublin or Cork, the more likely we will observe particularly severe cases in terms of violence, and we will probably be missing depositions from people who lost property but were not subject to violence.
Unlike contemporary data, where we can ensure representativeness of surveys, historical data and especially of the pre-modern period is likely far from ideal. We cannot go back in time to interrogate everyone, so we have to find a way to use the available data as much as possible. While we do not know every detail of how depositions were collected, we may make some reasonable guesses regarding potential biases to account for that in our analysis. We already know that the depositions do not represent everyone equally: neither by ethnicity and religion – mostly representing Protestants from England and to a lesser extent Wales and Scotland – nor by gender – by mostly representing testimonies of men or by social status, mostly representing the ‘middling sort’.
Luckily, there is a rich statistical toolkit to work with historical data, which can take sources of data selection (such as travel distances) into account. For example, we can employ regression analysis, a statistical method used to test whether there is a systematic relationship between the variables of interest in the data. Regression analysis allows us to account for additional factors, such as travel distances or other elements influencing travel. In other words, regression analysis allows us to compare the depositions from the areas that are the same distance from Dublin (or Cork). If some regions – for instance, areas that were recently colonised – were more prone to violence than the other places, then we would expect to see that those areas were indeed systematically different in terms of the incidence of violence. In this way, we can test different hypotheses numerically.
Using this and other methodologies, the VOICES project will provide a systematic overview of violence experienced by women during the 1641 Rebellion, while also considering potential biases in the data that may arise from selection mechanisms. By using the depositions as a source of numerical data, we aim to focus on women’s experiences during periods of social unrest, by quantifying the rates at which they fell victim to crimes, particularly sexual violence, across diverse geographical regions and social strata.
[1] 1641 Depositions project page: https://1641.tcd.ie/index.php/using-categories/