Literary criticism needs to explore how to meet the challenges posed by the digital humanities era. Photo: IC PHOTO

In his lecture titled “What are the Digital Humanities and Why are They Important?” Harvard Professor Peter K. Bol pointed out four important features that differentiate the digital humanities from traditional humanities disciplines: digital resources, data, applying computational methods to find and extract data, and using computational methods to analyze and visualize data to generate new information. These features encompass the complete research cycle—searching, collecting, organizing, and analyzing data—within a digital environment, where research methods of the digital humanities are applicable. Amid contemporary disciplinary integration and media convergence, the literariness defended by aestheticians has long been in jeopardy. Literary criticism needs to explore how to meet the challenges posed by the digital humanities era and how to effectively integrate important research methods in the digital humanities to facilitate innovation.

Integration of digital humanities methods

Digital humanities methods primarily involve text mining and data analysis, often leveraging “big data” or comprehensive data processing with relatively novel analytical and statistical methods. Engaging in macro-analysis of and research on literature that transcends the limits of time, space, and culture, digital humanities represent cross-disciplinary research and disciplinary integration. This has drastically altered criticism and research paradigms in traditional humanities disciplines. Research models such as database construction and visual analysis have undoubtedly facilitated scholars’ access to research materials. However, the rapid development of relevant research archives and database construction in recent years has caused frustration for “material addicts”—those devoted to exhaustive research. Rapid digitization and informatization necessitates a shift in criticism paradigms among those who possess traditional historical materials, while digitization and visualization of vast amounts of resources have rendered old concepts of criticism obsolete.

Research methods in the digital humanities have effectively broken down the outdated concepts and rigid paradigms of criticism. Moreover, their wide application has been advocated to establish a visual knowledge graph, thereby creating a relatively novel landscape of criticism and research. The project “Platform for Geographical and Chronological Information on Tang-Song Literature,” led by Professor Wang Zhaopeng from South-Central Minzu University and supported by the National Social Science Fund of China, is a good example of integrating traditional research on ancient Chinese literature and modern digital technology. This platform consists of two core elements. The first is basic data, which is to say the chronological and geographic information about writers’ activities and creations, extracted from three types of documents: chronologies of poets’ lives, annotated editions of their works, and related biographical studies for chronological and geographic data about Tang poetry. The second is the “vector maps,” a visual map platform established by vectorizing the Historical Atlas of China edited by Tan Qixiang and integrating it with literary research data.

This approach effectively addresses two major problems in ancient literary research—the dispersion of data, and time-space separation. Not only can it link knowledge across different disciplines and categories, it can also more intuitively integrate information such as when the writers were active and where their poems were written. This has not only renewed criticism and research, but also facilitated teaching and research.

Distant reading is an important approach in the digital humanities. Coined by Italian-American scholar Franco Moretti, distant reading is considered to be an important counterpart to the traditional approach of close reading. Contrary to what it may seem, distant reading is hardly mysterious. Characterized by large-scale text analysis, computational criticism, and collaborative reading, distant reading focuses on the overall perception of texts from a macro level, which is impossible to achieve with close reading. The “macroanalysis” approach proposed by Matthew Jockers, which aims to read and capture the common features of massive literary works, can also be regarded as a variant of distant reading as proposed by Moretti. Qualitative analysis supported by quantitative research, which complements the macroanalysis approach, advocates for the use of data-based means to reveal the changing trends and intrinsic laws of literary phenomena.

In fact, as early as the 1980s, Professor Chen Dakang already made an effort to unveil the enigma of authorship concerning Dream of the Red Chamber [an 18th-century Chinese novel written by Cao Xueqin, with uncertainty over the authorship of its final 40 chapters]. Without the aid of any artificial intelligence (AI) software, Chen carefully analyzed certain function words and idiomatic words and expressions in the novel through manual retrieval and mechanical comparison, coming to the convincing conclusion that the final 40 chapters of the novel were not written by Cao Xueqin. To some extent, this represents the organic combination of macroanalysis and qualitative analysis supported by quantitative research.

In fact, distant reading has already become quite common in contemporary criticism of popular online literature. As an “ultra-large text,” online literature differs vastly from traditional literature that is accustomed to elite reading. The lack of objective conditions for traditional close reading provides an opportunity for distant reading in the case of online literature. In recent years, with the aid of AI software like “One Leaf: Story Collection” (“Yiye Gushihui”) and other intelligent character relationship analysis software, researchers have achieved macro-statistical analysis of massive texts of online literature through exploring the distribution and visualization of connectors, as well as illustrating character relationships. Research methods in the digital humanities have helped reveal important facets of online literature, thus creating a new paradigm for criticism of online literature.

Research methods in the digital humanities, characterized by knowledge graphs, text mining, and data analysis, are accelerating their integration into traditional literary criticism, which has significantly expanded and enriched literary research and criticism. In addition, AI technologies, represented by data simulation and deep learning, are also penetrating into all aspects of literary research and criticism at an unprecedented rate. For example, the application of technologies such as sentiment analysis and intelligent mining is emerging, and AI applications featuring literary appreciation and dialogue generation based on deep learning are in full swing. This also offers opportunities and possibilities for the practice of intelligent criticism based on human-machine interaction and collaboration in the future. These advancements represent a gratifying change of the times.

Concerns over literary criticism amid digital humanities

The development of digital humanities technology can be a “double-edged sword.” Easier access to materials is not necessarily a boon for researchers on literary criticism. The vast amount of readily accessible research outcomes may also lead younger researchers to approach their work too casually or rely too heavily on established patterns. For example, with the rise of the “internet generation,” young students today often rely exclusively on thesis databases for research materials. However, much of what fills these databases consists of low-quality, repetitive publications, with few truly original insights. The author is convinced that most scholars share this experience: when their undergraduate students choose a topic for their graduation theses, certain stereotypical themes tend to reoccur frequently, such as “the tragic theme of Yu Hua’s novels” and “the female image in Eileen Chang’s novels,” which, in my opinion, is the enemy of contemporary academic research. This is precisely why Peking University Professor Chen Pingyuan lamented that our ability to search is strengthening, and yet our ability to think and read is weakening. It is necessary to note that this is also what worries literary criticism in the era of digital humanities.

Moretti’s distant reading represents multiple paradigm shifts in reading objects, subjects, and methods through its approaches of “second-hand reading,” large-scale reading, collaborative reading, and computational criticism, thus promoting the transformation of traditional literary criticism toward the digital humanities. These shifts are also believed to have effectively addressed the crucial issues of objects, subjects, and methods in literary criticism. Distant reading has revolutionized conventional reading modes and set a precedent for digital humanities literary criticism. However, as Nanjing University Professor Dan Hansong observed, “the significance of algorithmic criticism lies precisely in using AI to prompt critics to discover problems that have not been noticed before, and to help critics shed new light on texts.”

For literary criticism, research methods in the digital humanities advocate for the rapid extraction of elements from the text through distant reading, as it utilizes advanced technical means to provide new perspectives and tools for literary criticism. However, it does not and cannot completely replace the traditional approach of close reading. This is because technology can “extend the functions of human organs” and become a “proxy” for humanity, but it can never replace humans themselves. For Moretti, interpreting the findings of distant reading remains reliant on scholars’ knowledge and understanding shaped by a close reading of the text. In short, the in-depth analysis of a work is always inseparable from the unique emotional recognition of human beings, which makes human reading irreplaceable.

It is in this sense that the more rapidly digital humanities technology develops, the more necessary it is to keep alert to its risks. In other words, in the era of digital humanities, humanists’ vigilance and worries will never be absent. We need to be alert to the risk of slipping from traditional “text fetishism” to modern “technological fetishism” in the era of digital humanities, and also alert to people’s obsession with machines due to objective circumstances. We should also be concerned with the unequal access to cultural capital that may have been further exacerbated by the differences in information access caused by the digital divide. At the same time, as scholars have repeatedly called for, we should not blindly embrace the rise of digital humanities. Even if digitization is growing exponentially, we should continue to seek out real documents. Literary criticism must determine how to best preserve the beauty of literary imagery and emotion, while also safeguarding the last garrison that defines us as humans in the era of digital humanities.

Xu Gang is a research fellow from the Institute of Literature at the Chinese Academy of Social Sciences.