As a scholar of English interested in Children’s Literature, I have spent hours building data scrapers, contemplating user experience, exploring information design, and developing scripts to explore literary and visual content for distant reading of multimodal print artifacts. When confronted with problems of tool and method, I went so far as to form a team of scholars to build TANDEM, a tool for generating data from text and images. Built on Django, written in Python, it utilizes OCR, natural language processing, and computer vision to build a useful entry level dataset for a scholar exploring images and text.
Picturebooks are the primary focus of my recent work. In Children’s Picturebooks: The Art of Visual Storytelling, Martin Salisbury and Morag Styles stake, “Today’s picturebook is defined by its particular use of sequential imagery, usually in tandem with a small number of words, to convey meaning.” A picturebook is a visual story told in multimedia. It is images and words. It cannot be one or the other. It is both. Therefore, exploring them using quantitative methods that we can translate into visual stories in themselves via data visualization and corpus manipulation seems like an obvious endeavor for a modern scholar.
The Oxford English Dictionary (OED) defines “humanities” as:
The branch of learning concerned with human culture; the academic subjects collectively comprising this branch of learning, as history, literature, ancient and modern languages, law, philosophy, art, and music.
A footnote continues the definition, stating:
The humanities are typically distinguished from the social sciences in having a significant historical element, in the use of interpretation of texts and artefacts rather than experimental and quantitative methods, and in having an idiographic rather than nomothetic character.
The OED substantiates the perceived polarization of academic techniques. Humanities scholars use interpretation, not quantitative methods. This is approach to the academic inquiry is antiquated. Many people before me have addressed this issue, and there is plenty on to read on both sides of the topic.
We can pinpoint many reasons why humanities scholars are utilizing quantitative and computational methods to study their material. Perhaps it is the availability of tools and tutorials. Perhaps it is a venture based in paranoia as STEM and the social sciences are becoming focal points in the institution, leaving English and Philosophy undergraduates to be balked at. Again, a topic that is ripe with discussion at all levels of academic study. What curriculum is worthy? How should things be done? What topics and techniques are worthy of pride, and more importantly tenure.
Children’s lit scholars exist in a unique niche. Karen Coats opens “Fish Stories: Teaching Children’s Literature in a Postmodern World” with:
Teaching children’s literature in a university English department is an enterprise
fraught with personal and professional risk. No matter how sophisticated
your theoretical commitments, no matter how learned you are in and
beyond your subject area, you suffer the bemused and patronizing smiles of
peers who find the aesthetic virtues of Dr. Seuss less worthy of study than
those of, say, Thomas Hardy or Emily Dickinson.
The study of picturebooks is much like the emergence of digital humanities in academia. To hardened computer scientists, digital humanities scholars are the subject of bemusement and patronizing smiles. Someone recently said, “You can write baby code. You need a computer scientist to build the powerful stuff.” At this stage, I think I can agree with that. The problem is, I want to build the powerful stuff. The solution is, I need to initiate conversations with those who understand my theoretical, methodological, and ultimately computational problem to get to my answers. Software may generate data, but unless I have at least a cursory understanding of how that data is generated, I would be foolish to blindly take it as truth.