Introduction

Literature creates imagined worlds, and we never quite know whether they are created to make us feel at home or set us off by changing the rules of how things are. Literature is also a way of expressing subjective impressions of the world (and the self) in ways that are not intuitive.

Therefore, the question of sensation and the perception of phenomena are important elements of understanding texts, although they are often under-reflected, and not central to analysis.

It is so self-evident that the textual representation of sensation is important that one quickly moves away from it. Like style, sensation may be difficult to pinpoint, and thus formalise. On the one hand, it is possible to scan texts for certain words that imply subjectivity and sensation, and hence to characterize this. This may be useful, but on the other hand, many texts may also use less obvious ways to describe such elements.

An example of good use of computational approaches was one carried out by a group of students that explored Franz Kafka's The Trial. Their hypothesis was that they could show how law, punishment, and guilt permeated the text, and although these phenomena were present, they found that walls, doors, and rooms carried more weight. A thematic approach led the group to focus more on a particular way of sensing the world that may not be surprising, but they could confidently describe how theme and sensation worked together. Thus, a computational approach to the study of sensation could be carried out with a number of the methods described for the previous topics, but with a stronger focus on this specific element of world-making.

Applications

Elementary

One way to computationally approach sensation is to search a text or a corpus for words distinctly associated with sensation, subjectivity, and emotional expression. This allows you to get an overview of these words and observe them in their contexts, both of which can be useful when characterizing sensation and understanding the text in which it appears. This approach can be simple and easy if applied to only one short text or passage, but it is not trivial to apply it to e.g. a whole novel or a big corpora of texts. It can be difficult to determine which words to search for, as the words you could expect to find are not always obvious and rarely known in advance. One starting point can be using a reverse dictionary (e.g. here or here) and searching for “sensation” or words believed to be specifically associated with sensation extracted from lists of words observed in other contexts. Use Concordle to start exploring in which contexts words for sensation appear, or create your own annotation project for sensation analysis using e.g. Digital Mappa or Annotation Studio on a work of your choice.

Advanced

Although sensation and perception might seem hard to grasp computationally, it is possible to approach these concepts, for example by identifying words that are high or low in sensation (using for example the resources above). By statistically analysing the proportion, occurrence, and significance of these words in a text, it is possible to get insights into the sensory phenomena and impressions the text invokes.

Get inspired by the Stanford Literary Lab’s Pamphlet 7 (Holst Katsma, 2014), which presents an investigation of loudness in the novel. Following Katsma, novels are full of voices, and by establishing a “grammar of loudness”, it is explored how “loud” the reading experience appears in a story. Especially dialogues might vary in loudness, characterised by expressive punctuation, and verbs such as “cried”, “shouted,” or “screamed”. Apply the approach on a novel of your choice and reflect on your findings and what they reveal about the characters and the atmosphere in the story. Can you identify other similar sensation vocabularies to trace other senses such as taste or smell? Have a look at David McClure’s blog post and try plotting the “sensation” topic you’ve defined on a novel or a corpus.

Resources

Scripts and sites

• Annotation Studio, a collection of web-based and collaborative annotation tools.

• Digital Mappa, an open-source digital humanities platform for open-access workspaces, projects, and publications.

• Concordle, a word cloud and concordance generator.

• FuD, an annotation tool by Trier University.

• David McClure’s GitHub repository for literary interior, i.e. how words distribute in a text, and for literary interior analysis.

Articles

• Acerbi, A., Lampos, V., Garnett, P., & Bentley, R. A. (2013). The Expression of Emotions in 20th Century Books. PLoS ONE, 8(3), e59030. https://doi.org/10.1371/journal.pone.0059030
• Brand, C., Acerbi, A., & Mesoudi, A. (2019). Cultural evolution of emotional expression in 50 years of song lyrics. Evolutionary Human Sciences, 1, E11. http://dx.doi.org/10.1017/ehs.2019.11 

• Katsma, H. (2014). Loudness in the Novel. Stanford Literary Lab Pamphlet 7. https://litlab.stanford.edu/LiteraryLabPamphlet7.pdf 

• McClure, David. (2017, July 31). A hierarchical cluster of words across narrative time. Stanford Literary Lab. https://litlab.stanford.edu/hierarchical-cluster-across-narrative-time/
   

Rhythm »