This page is referenced by:
Computational Literary Studies: Underused and Underrated
The merit of the Computational Literary Studies is often questioned due to its shallow, yet practical nature. In this essay, I will prove that although CLS cannot analyze texts in as much depth as humans, it can provide useful analysis of historical trends.
In the Digital Humanities, there is a subfield that works with the application of digitized analysis of a manifold of texts and literary pieces. This subfield is commonly referred to as Computational Literary Studies (CLS), which uses computer software to analyze a multitude of texts or books, where its conclusion is a generalized summary of the trends from the texts. The conclusions are then incorporated into our original understanding of texts and trends. The CLS examination of word frequency allows inspection in aspects of historical or contemporary texts; such as the evolution of gender in fictional writing.
CLS is often criticized for producing insignificant or inaccurate results. In Nan Z. Da’s “Digital Humanities Debacle” she states that “Literary objects are too few, and too complex, to respond interestingly to computational interpretation — not mathematically complex, but complex with respect to meaning, which is in turn activated by the quality of thought, experience, and writing that attends it” (2019). She believes that CLS cannot conclude useful information from literary texts because text requires human interpretation for analysis. In Da’s view, mathematical complexity is simple compared to the complexity of literature. This is because numbers don’t entail elaborate details such as human emotions. CLS primarily uses statistical data to draw conclusions, therefore is commonly seen as inconsequential due to the depth that literature requires to be thoroughly understood. In this paper, I will prove that even though some scholars believe that CLS lacks in-depth analysis, the CLS examination of word frequency can conclude insightful trends of historical texts; such as the evolution of gender in fictional writing.
A common belief is that CLS is incapable of presenting specific and detailed conclusions. In Da’s “The Computational Case against Computational Literary Studies”, she disputes CLS’s accuracy of observation of historical, cultural, and literary texts. She believes that the conclusions and analysis made by the CLS lack critical analysis and dulls the understanding of culture.
“In CLS data work there are decisions made about which words or punctuations to count and decisions made about how to represent those counts. That is all” (Da 606).
Here, the fallacy that the analysis of word count can represent human emotion or culture is addressed. She thinks that the magnitude of CLS ends at the conclusion of statistical data and it cannot be favored over human analysis. Nan Da makes a thoughtful point in showing that CLS only aims to answer one of six questions; which include, “aboutness, influence, relatedness, connectedness, generic coherence, and change over time all represented by the same things” (605). CLS is not able to give enough analysis for it to replace human interpretation and analysis of texts. Its computation is still limited in the ability to draw abstract or even specific conclusions, but this does not render CLS futile.
CLS is able to derive cultural changes and trends by evaluating a corpora of texts. In Cultural Analytics, “The Transformation of Gender in English-Language Fiction”, Ted Underwood, David Bamman, and Sabrina Lee collectively refute the common conception that CLS is morally insulting to literature and unequivocally inaccurate. By inspecting the evolution of gender in fictional texts with CLS, they were able to examine the change in prominence of characterization from the 1900s to the 2000s (2018). According to the authors’ research, the distinction in the language used in fictional books to describe male and female characters were more easily distinguishable in the nineteenth century than in contemporary texts (2018). They state that before the mid-nineteenth century, there was a clear distinction between men and women in the words used to describe and relate to each of them. With the progression of time, these ‘gender assigned words’ started to become blurred and obsolete(2018). CLS also allowed them to see the trend that there was a negative relationship between time and amount of female writers in fiction. As time passed, the number of females writing in fiction decreases significantly(2018).
Despite CLS’s ability to track trends and ideas from history, CLS is still commonly criticized for its lack of ability to do in-depth critical analysis. Da advocates for this idea in doubting the abilities of CLS. “Because of the way the data is treated, CLS can make macrohistorical claims that are statistically uninformative” (Da 610). Her criticism is that CLS is only able to make macrohistorical claims, which entails general trends such as social movements. This is consistent with research as CLS has yet to prove that it is capable of explaining or representing specific historical events, such as the extension of property rights to women in 1849 (Women’s National History Museum 1). From CLS, we are able to see trends in history, but we are not able to see the specific affairs within these trends. Da is correct in supporting the argument that CLS is not able to conclude critical and precise information from text, but this should not belittle its ability to show significant historical trends. Da’s premise is accurate on the surface, but the belief that these observations are trivial is a misjudgment in her argument.
Computational analysis amplifies the human understanding of literature and summarizes parts of texts to help us understand historical trends. Underwood, Bamman, and Lee derive most of their argument off of the computational analysis’ ability to detect the “masculinization of fiction” in literature from the 18th to the 20th century(11). “When we separate authors by gender, we find that women were becoming less prominent even in books by women across this century” (11). In the authors’ analysis, they identify the gender regressive movement in fiction in the past 200 years. The use of CLS was able to detect a drop from 1 out of every 2 books of fiction being written by females in 1870, dropping to 1 out of every 4 books in 1970 (10). This declination of published female writers was able to be identified by the computational analysis of different words that were used exclusively for specific genders, and by examining the use of these words in hundreds of different texts. They also state in the article that “We are confronted with a paradoxical pattern. While gender roles were becoming more flexible, the space actually allotted to (real, and fictional) women on the shelves of libraries was contracting sharply” (13). CLS is able to show us aspects of culture and change of societal standards. The Women’s Rights movement began in 1849. From this point onwards, the line between duties, descriptions and roles of men and women slowly began to disappear. By analyzing history and the findings made by Underwood, Bamman, and Lee, a parallel is able to be drawn in the two trends. Thus, showing the ability for CLS to be able to show shallow, yet practical information about culture and history.
With the analysis of texts, CLS has been able to progress and enhance the practice of literary analysis. It observes repetition and commonality of words and phrases, which allows the conclusion to consist of data about gender and culture. It currently lacks the faculty to provide in-depth, critical analysis of texts; but CLS has provided the world with notable information about historical trends and tendencies in texts.
Da, Nan Z. “Computational Methods Repeatedly Come up Short.” The Chronicle of Higher Education, 27 Mar. 2019, https://www.chronicle.com/article/The-Digital-Humanities-Debacle/245986.
Da, Nan Z. “The Computational Case against Computational Literary Studies,” Critical Inquiry. Volume 45, Issue 3, University of Chicago Press, March 2019, pp. 601-639
Underwood, Ted, David Bamman, and Sabrina Lee. “The Transformation of Gender in English-Language Fiction.” CA: Journal of Cultural Analytics, February 13, 2018. http://culturalanalytics.org/2018/02/the-transformation-of-gender-in-english-language-fiction/.
Ted Underwood, David Bamman, and Sabrina Lee, "The Transformation of Gender in English-Language Fiction," Journal of Cultural Analytics. Feb. 13, 2018. DOI: 10.31235/osf.io/fr9bk
“Woman's Suffrage Timeline.” National Women's History Museum, https://www.womenshistory.org/resources/timeline/womans-suffrage-timeline.