Data Critique: Trans-Atlantic Slave Trade Voyages Database
by Tyson Luneau and Kristian Price
The Trans-Atlantic Slave Trade Voyages Database includes records of 36,002 voyages which engaged in the sale and purchase of African slaves. The recorded voyages in the database begin as early as 1514, with the last voyage occurring in 1866. There is no one distinctive, geographic pattern of these voyages, with ships traveling between Africa, Europe, and the Americas, occasionally not traveling between continents but rather from colony to colony, avoiding the middle passage altogether. Some of the voyages identified did not even contain slaves on the ship, as the captain and crew simply returned to their original ports. While some of these instances could be considered statistical outliers, it nonetheless reflects the diversity of data within the dataset.
The sourcing of this data was conducted primarily by compiling legal and economic records from individual ports, as well as advertising contained within newspapers. These records were initially created for the purposes of insurance, legal protection, and duties. Early productions of the dataset only contained information from English, French, and Dutch sources, thereby limiting the scope of the transatlantic slave trade. However, between 2001 and 2005, sources from Spanish and Portuguese voyages added the element of Latin American slave trading expeditions to the dataset, giving a slightly more comprehensive view of the trade network.
The dataset contains records of voyages that vary significantly in detail. Nearly all records include the name of the ship, captain, year of arrival, points of origin, slave purchase location, fate of the voyage, and the number of slaves embarking and disembarking. However, there are other categories where data is missing more frequently, due to intentional or unintentional omission. Furthermore, the existence of statistical outliers in the number of slaves present can dramatically swing figures based upon that data, such as the mortality rates. For example, voyage 3210 contained only two slaves on the ship, one of whom died en route to the destination, resulting in a mortality rate of 50%. Figures such as these skew the average mortality rate significantly, thereby opening up the potential for misinterpretation of the data.
Other limitations include the lack of context and the quantification of information associated with human tragedy. The lack of qualitative information on the slave trade can lead a less-informed viewer to interpret the slave trade as a purely economic interaction, omitting the experiences of those subjected to the Middle Passage. Additionally, this dataset does not provide any information regarding the journey of slaves after the initial disembarkment en route to the final destination.
Not inherently the fault of the compilers, this dataset does not necessarily represent the full scale of the transatlantic slave trade. One of the most jarring dynamics, not uncommon among datasets of such a substantial size, is the omission of data in a large number of records. Historians of the slave trade would suggest that some of this data was intentionally omitted or skewed, such as slave deaths or instances of resistance, often as a means of avoiding lawsuits or other legal ramifications. Other records have data that was unintentionally omitted, such as the length of the Middle Passage or tonnage of a ship, due to differences in record keeping spatially, temporally, or between different legal systems. The wide array of categories included in the data set is due to differences in record-keeping, where some ports and companies were far more elaborate in their practice than others.