Dr. Jonathan Y. Cheng from the University of Nebraska, Lincoln, analyzed 13,000 novels published from 1850-2000 and found that more and more descriptions of characters in novels focus on body parts. In 1850, male characters were described with body parts about 7% of the time, compared to 15% in 2000. Female characters’ body part descriptions rose from 12% to 17%. 

To further analyze these novels, Cheng made his computer guess a character’s gender based on specific words used to describe body parts in a novel. After this, he divided the 13,000 novels into groups based on whether the book was written by a male or female author. He wanted to see if grouping them in this way would make a difference in the ability to guess the character’s gender as it is described with body parts. As a result, the analysis showed that it was much easier to determine the gender of the character when the author was male than when the author was female. 

Cheng is not only making a case for current authors to break away from using body parts to describe women so often, but also encouraging authors to embrace new ways to discuss fictional characters through gender fluid terms and actions. Moving away from this traditional way of describing characters will also allow more readers to feel accepted and seen through novels if they make this change. Researching the frequent overuse of body parts to describe women, as Cheng does within this article, will hopefully bring awareness to this issue and allow both current and future authors to do better than authors of the past.

Full Article: https://culturalanalytics.org/article/11652-fleshing-out-models-of-gender-in-english-language-novels-1850-2000

Six graphs from the article demonstrate trends over time for "potato," "cinema," "cigarette," "couch," "radio," and "sweater"

Researchers from the Netherlands, China, and Denmark analyzed 18.6 million advertisements and 11.5 million articles from two Dutch newspapers published between 1890 and 1989. They had several major research questions. Do ads shape articles, or the other way around? Are there measurable trends in how the articles and ads use language? And how are trends different for different kinds of products?

Their analysis found that advertisements do shape articles–but it also works the other way around. Trends in words used in 20% of the advertisements predicted words in future articles. On the other hand, about 17% of the articles predicted words used in future advertisements. But 49% of the trends seemed to be driven by outside forces.

Trends in ads were much shorter-lived than trends in articles. Words like “living room,” “couch,” and “lamp” were persistent in articles but not ads. Both ads and articles persistently featured the words “cigarettes,” “heels,” and “cauliflower.” But technological words like “film” and “radio” were short-lived, not persistent in either ads or articles.


Melvin Wevers, Digital Humanities Lab, KNAW Humanities Cluster, Netherlands
Jianbo Gao, Center for Geodata and Analysis, Faculty of Geographical Science, Beijing Normal University, Beijing, China
Kirstoffer L. Nielbo, Center for Humanities Computing Aarhus, Aarhus University, Denmark

Full article at Digital Humanities Quarterly:


A recent study that analyzed the faces within 3,389 issues of Time magazine published between the years of 1923 and 2014 found that the majority of faces represented belonged to white men. As one of America’s most notable news magazines since 1923, Time magazine demonstrates how popular attitudes within America have evolved over time. The study used three datasets. The first dataset used a machine to determine the gender of the face based on samples from the magazine. The second dataset utilized human labor to classify the faces based upon several factors like race, gender, or age. The final dataset included the raw data used to create the second dataset. The researchers plan to use the data from the study in the future to determine how the representation of women within the magazine has evolved over time, particularly in advertising purposes.

Full article: https://culturalanalytics.org/article/12265-faces-extracted-from-time-magazine-1923-2014

The Time magazine project: https://magazineproject.org/TIME/

The project’s open dataaset: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/JMFQT7

A photograph and a digital model of dunes as they exist today and as they existed 200 years ago
A destroyed dune known as “Cossack hill,” as it exists now on the left, and below right, a 3D model of how it looked 200 years ago.

New digital mapping research reveals that the Toruń dunes in Poland once rose 45 meters high, began to form 13,900 years ago, and remained virtually unchanged until the second half of the 20th century. In the last 200 years, dunes have shrunk 26.5% in the city limits, and 60.2% in the most urbanized part of the city. The greatest part of this degradation happened after the late 1950s. To make this comparison, researcher Paweł Molewski of Copernicus University in Poland compared maps and drawings of the landscape from 1793-2000, and after 1965 was able to use satellite views also. He used a coordinate system, the Topographic Object Database, and software including ESRI ArcGIS 10.4.1 and Global Mapper 17 to map, and Voxler 3 to model the dunes. Ultimately, Molewski’s work suggests that if Toruń residents want to preserve what dunes remain, their further urbanization and development will have to protect the natural features more deliberately.


A recent blog written by Ewan McAndrew, a Wikimedian in Residence at the University of Edinburgh, declares that students from different disciplines and backgrounds can help close the knowledge and representation gap by contributing their scholarship to Wikipedia. McAndrew claims that this limitless potential is enticing for educators with diverse student populations across the world and many incorporate it into their classrooms to meet their students’ specific interests. He claims that for so long, many educators have wondered how extensive the link between Wikipedia and the University should be. Wikipedia has widely been used in education to warn students of open-access information systems where knowledge can be right, wrong, or even missing. McAndrew argues that educators should instead teach their students how to contribute their own knowledge to Wikipedia, stating that this scholarship will encourage other students and academics to view Wikipedia as an effective learning technology that fosters a better web and world.

See the full article here: https://blog.wikimedia.org.uk/2020/02/knowledge-activism-vs-passive-consumption-rethinking-wikipedia-in-education/

A recent study of use of the word “Cherokee” on Twitter shows that stereotypes of Natives outnumber positive mentions by three to one: Susana at Indigenous Engineering found that words like “tomahawk” and “Pocahottie” were far more common than mentions of tribal sovereignty and citizenship. The study marks a sad anniversary for natives, the February signing of the Dawes act, which broke up tribal lands and resulted in millions of acres being seized by non-native white settlers. This year, February also included several major Democratic presidential primary debates, which took up discussion of Sen. Elizabeth Warren’s contested claims to Cherokee ancestry. Susana’s analysis considers the use of the word “Cherokee” before and after the  February 7th debate, concluding: “It should be noted that on February 8th, 2020, the 133rd anniversary of the signing of the Dawes Act, Cherokee identity was again demonstrably hijacked by a white politician–this time not thought direct legislation, but rather words & actions in the social sphere.”

See full data and analysis on the Indigenous Engineering website: https://indigenous.engineering/projects/Cherokee-Online.html

0 Comments/by