When contemplating the question of data in humanities research, I begin to think about the size or scale. Since the humanities deals with what makes us uniquely human, I begin on a personal basis. As an individual, if I have a problem or perhaps on the more positive swing, try to determine a point of personal interest, I start with facts, break down the process, consider environmental influences, and acknowledge results. While this type of problem-solving includes a psychological nature, which is often hard to measure, many of the tools and processes utilized share a commonality with its scientific cousin. Hence, the practice or concept of sharing tools and processes between disciplines and incorporating appropriate modifications and/or additional components has existed for generations. Therefore, the concept of humanities or even specifically digital humanities (the study of the humanities integrated with the tools used by STEM), utilization of data sets, management tools, and forms of research, is a logical evolution of the field.
The utilization of data in the humanities is one of the foundational components of research in the discipline. Oftentimes, the question to be answered by the research stems from looking for a pattern, answering a curiosity, solving a problem, or searching for a greater understanding of issues, timelines, influences, or cultural nature. Because the data is a fundamental component of this process, it must be as accurate and reflective of the truth as possible. Hence, the importance of clean data.
Issues with limited, unreliable, or messy data can be one of the larger contributing factors to inaccurate results, as the information utilized in the study is one of the major cornerstones of a research process. Hadley Wickham, author of “Tidy Data,” states that 80% of data analysis is spent on the process of cleaning and preparing the data. Through the study of this and the additional weekly readings, it is apparent that there are pros and cons of how it is conducted, as what might be apples to apples and oranges to oranges in a more scientific realm, might not translate so well into the humanities, as the variations between the subjects are not always cut and dry. It is these variations that can lead to unintentional misrepresentation and, therefore, inaccurate results.
Katie Rawson and Trevor Munoz’s article “Against Cleaning” brings up an interesting concept that applying traditional methods of cleaning data used in other fields might actually suppress diversity in the field of humanities studies. Other points to consider were gleaned by reading Catherine D’Ignzio and Lauren Klein’s “The Power Chapter,” which discussed algorithms that may unintentionally be based on a limited model scope. Other questions that came to mind when discussing this same article are the disparities between races in the maternal health crisis, including the subjectivity of terminology, no uniform way of tracking information for studying the issue, and a lack of representation of information from other healthcare and maternal delivery alternatives, such as home births and the use of doulas and midwives.
While the importance of research in Digital Humanities continues to grow as we see an evolving scope never experienced before, the tools and methods we utilize must remain tailored to the specific nuances of our discipline to ensure the accuracy of our findings and preservation of the landscape that is uniquely ours as human beings.
Jamie– I completely agree that size and scale are very important concepts to data work! As we live and work in an age of “big data”, I often have to remind myself that not all data is “big.” Anything can be data (or contain data within it). As a historian, I think there are certain challenges when approaching sources as “data” and we have a responsibility to avoid the pitfalls of quantitative research. In some cases, data-driven history can mean a reduction or simplification of the narrative, or replicating harm.
While data can bring new perspectives to our work, it’s always important to remind ourselves of placing that data in the context of its creation and in the context of a qualitative reading of the sources!