Modeling Bodies in Romantic Texts and Scientific Culture

Historians of science and medicine recognize the British Romantic era—during which anatomical study flourished and physiology developed—as the dawn of professionalized medical science.¹ Language about bodies, from newspapers to fiction and poetry, contains data that can be studied and modeled to help map the history of our bodies as objects of medical science. If, as William Abberley writes, a primary function of Romantic-era replicas of bodies was to order and make sense of them, to turn them into ideal types, we might argue that such a trend continues today; we structure and generalize data collected about diverse bodies to treat them prescriptively.² How can we look to nineteenth-century texts to better understand the origins of this trend?

Building on histories of gendered bodies by Deborah Lupton and Richard C. Sha, future Romantic literary scholarship could look more closely at how women and writers of color represented bodies in their poetry and prose—close studies of their semantic, syntactic, and narrative choices could produce models for understanding the intersection of scientific medical science and subjective experiences of bodies.³ Algorithms that explore semantic similarities could be used as points of departure to understand how literary descriptions of bodies reflect social dynamics, aesthetic values, and factual knowledge.⁴ Approaching this problem with computational and analytical tools could help answer questions such as: How did male and female Romantic authors describe and produce their own images of bodies? Did marginalized authors produce alternate representations of bodies? For instance, how are bodies integral to the narrative in The History of Mary Prince? How does Charlotte Smith refer to her body in Elegiac Sonnets? How are bodily rituals depicted in Emma Roberts's Oriental Scenes, Sketches, and Tales? Which other texts can we discover by looking for patterns of meaning that we identify in these works?⁵

I'm excited by the prospect of scholarship that takes advantage of newly available digital texts by under-represented Romantic-era authors and open-source machine learning tools to model semantic relationships within them.⁶ These models might reflect a broader experience of Romantic culture and, in turn, alter how we understand the evaluation and treatment of bodies, diversely construed, in today's scientific climate. As Foucault argues, power becomes disciplinary when it regulates biological practices such as reproduction and death.⁷ With Foucault's ideas in mind, computationally driven Romantic literary scholars can use learning representation techniques⁸ to generate language models that capture non-normative somatic experiences. These models can be used to augment feminist and postcolonial epistemologies of medical science and emphasize the political dimensions of early scientific agendas.

Canon expansion, for this type of project, is an act of data democratization, which, as a broader trend, has the potential to impact evolving medical practices that increasingly blend decision-making, technology, and data-driven diagnostics. When humanities scholars create new methods to represent information diachronically, they are creating tools to revise canonicity and narratives of history to be more global and inclusive. Those revisions can and should be applied in other domains.