From Close Reading to Digital Analysis: A Reflective Journey through Digital Humanities

From Close Reading to Digital Analysis: A Reflective Journey through Digital Humanities

Introduction

This blog is written as a task assigned by the Head of the Department of English (MKBU), Prof. and Dr. Dilip Barad Sir, as part of the Digital Humanities course. The aim of this blog is to document my learning experience while exploring various Digital Humanities tools and activities, including Computer-Generated Poetry, CLiC (Corpus Linguistics in Context), and Voyant Tools.

These activities introduced me to innovative methods of literary analysis by combining traditional close reading with computational approaches such as corpus analysis, text mining, and data visualization. Through hands-on exploration of these tools, I gained a deeper understanding of how technology can assist in interpreting literary texts, identifying textual patterns, and examining concepts such as authorship, creativity, language, and cultural context. This blog presents my experiences, reflections, and learning outcomes from each activity, highlighting the growing significance of Digital Humanities in contemporary literary studies.

Here is the link to the professor's blog for background reading:  Click here

What if Machines Write Poems?

A Digital Humanities Exploration of Computer-Generated Poetry

Introduction

The emergence of Artificial Intelligence (AI) has transformed almost every field of human knowledge, including literature. Poetry, traditionally regarded as the highest expression of human imagination and emotion, is no longer created exclusively by human poets. Today, computer algorithms and AI language models are capable of generating poems that imitate human language, imagery, rhythm, and even emotional expression. This Digital Humanities activity, based on Dr. Dilip Barad's blog "What if Machines Write Poems?", encouraged students to critically examine the relationship between technology and creativity by engaging with AI-generated poetry and identifying whether selected poems were written by humans or machines. The activity challenged conventional notions of authorship, originality, and literary creativity.

Objectives of the Activity

The primary objectives of this activity were:

• To understand the concept of Computer Poetry and Generative Literature.
• To explore AI-based poetry generation tools.
• To examine the role of Artificial Intelligence in literary creation.
• To compare human-authored poems with machine-generated poems.
• To reflect upon changing definitions of creativity and authorship in the digital age.
• To understand the significance of Digital Humanities in contemporary literary studies.

Activity 1: Human or Computer? – Google Form Quiz

Can a computer write poetry? | Oscar Schwartz

The next task involved completing a Google Form titled "Poem Written by a Human or a Computer?" In this exercise, several poems were presented without revealing their authorship, and participants had to determine whether each poem had been written by a human or generated by a computer.

The experience proved intellectually stimulating because many machine-generated poems successfully imitated human poetic expression. Their use of metaphor, imagery, symbolism, and emotional vocabulary often made them appear convincingly human. Conversely, certain human poems appeared deceptively mechanical due to their experimental style.

This activity demonstrated that readers frequently rely on subjective assumptions while evaluating literary authenticity.

Activity 2 : NPR "Human or Machine" Poetry Challenge

I also participated in the NPR interactive quiz titled "Human or Machine: Can You Tell Who Wrote These Poems?" The activity presented six sonnets, each of which had to be identified as either human-written or machine-generated.

Although I correctly identified four poems, two responses were incorrect because the machine-generated poems displayed convincing poetic qualities, including coherent imagery, emotional vocabulary, and formal structure. Likewise, one human-authored poem seemed artificial because of its unconventional diction and fragmented imagery.

The activity clearly illustrated that computational poetry has reached a stage where readers often struggle to distinguish between human creativity and algorithmic production.

Critical Observations

Several important observations emerged from this activity.

Firstly, AI-generated poems have become increasingly sophisticated in terms of vocabulary, syntax, rhythm, and imagery. Machines successfully imitate poetic conventions by analysing vast literary corpora, enabling them to produce structurally convincing poems.

Secondly, despite their linguistic competence, many AI-generated poems occasionally reveal semantic inconsistencies or emotional discontinuities. While the language appears poetic, deeper experiential meaning is often absent.

Thirdly, human poetry generally reflects lived experience, cultural memory, historical consciousness, and emotional authenticity. These qualities remain difficult for machines to replicate fully because AI generates language statistically rather than through personal experience.

Finally, the activity demonstrated that readers themselves frequently judge poetry based on stylistic expectations rather than actual authorship, making human and machine poetry increasingly difficult to differentiate.

Reflection

Before completing this activity, I believed that poetry represented one of the few creative domains beyond the reach of artificial intelligence. However, interacting with various AI poetry generators and completing the Human-or-Machine quizzes significantly altered my perspective. I realised that computational systems are capable of producing poems that imitate many features traditionally associated with human creativity.

Nevertheless, I also recognised an important distinction between imitation and experience. AI can reproduce patterns of emotional language, but it does not possess consciousness, memory, or lived emotions. Human poets transform personal experiences into artistic expression, whereas machines generate statistically probable linguistic sequences based on learned data.

Therefore, rather than replacing poets, Artificial Intelligence should be understood as a collaborative creative tool that expands possibilities for literary experimentation while simultaneously encouraging readers to reconsider established ideas of authorship and creativity.

Learning Outcomes

This Digital Humanities activity provided me with valuable academic and practical insights into the relationship between literature and artificial intelligence. I developed a clear understanding of the concepts of Computer Poetry and Generative Literature, and learned how Artificial Intelligence, through Natural Language Processing (NLP) and Machine Learning (ML), can generate poetic texts that imitate human creativity. By experimenting with different AI-based poetry generators and participating in Human or Machine? quizzes, I observed how computational systems can produce poems with convincing imagery, rhythm, and emotional language. This experience highlighted the increasing difficulty of distinguishing between human-authored and machine-generated poetry and challenged my traditional understanding of creativity and authorship.

Furthermore, the activity deepened my appreciation of Digital Humanities as an interdisciplinary field that integrates literary criticism with computational technologies. It encouraged me to critically reflect on evolving concepts of originality, creativity, and the role of the author in the age of artificial intelligence. Most importantly, I realised that while AI can successfully imitate poetic forms and linguistic patterns, meaningful literary interpretation remains a fundamentally human endeavour. This activity strengthened my analytical skills, broadened my perspective on contemporary literary studies, and inspired me to explore innovative digital methods alongside traditional approaches to literary research.

Conclusion

This activity offered valuable insight into the evolving relationship between literature and artificial intelligence. It demonstrated that machines are increasingly capable of generating poems that resemble human literary expression, thereby challenging traditional assumptions about creativity and authorship. Although AI can effectively imitate poetic language and stylistic patterns, human poetry continues to derive its uniqueness from lived experience, emotional authenticity, and cultural consciousness.

From the perspective of Digital Humanities, machine-generated poetry should not be viewed as a threat to literature but as an opportunity to rethink literary production, interpretation, and creativity in the twenty-first century. As Artificial Intelligence continues to evolve, the future of poetry may increasingly involve collaboration between human imagination and computational intelligence, opening new possibilities for literary innovation.

The Fireplace Pose and Cultural Meanings: Exploring Victorian Fiction through CLiC

A Digital Humanities Experience with Corpus Stylistics

Introduction

Digital Humanities has transformed the way literature is studied by combining traditional literary interpretation with computational methods. Instead of relying solely on close reading, researchers can now analyse large collections of literary texts using corpus-based tools that reveal recurring linguistic patterns, themes, and cultural practices. One such powerful platform is CLiC (Corpus Linguistics in Context), developed by the University of Birmingham.

For this laboratory activity, I explored Activity 13: "The Fireplace Pose – Texts and Cultural Context." The activity examines how the seemingly ordinary image of standing near a fireplace in nineteenth-century novels reflects broader social values, gender roles, and power relations. Through concordance searches and keyword analysis, I investigated how Charles Dickens and other Victorian novelists repeatedly describe characters in relation to the fireplace, revealing that literary language often encodes cultural meanings beyond the surface of the text.

What is CLiC?

CLiC (Corpus Linguistics in Context) is a web-based corpus analysis tool designed primarily for literary studies. It enables researchers and students to examine large collections of literary texts using corpus linguistic techniques such as concordance analysis, keyword searches, collocation analysis, and KWIC (Key Word in Context) displays.

Unlike traditional close reading, CLiC allows readers to observe recurring textual patterns across entire novels rather than isolated passages. This combination of distant reading and close reading provides deeper insight into an author's stylistic choices, narrative techniques, and cultural representations.

For postgraduate students of literature, CLiC demonstrates how computational tools can enrich literary criticism without replacing interpretative reading.

Aim of the Activity

The objective of this activity was to investigate how the word "fire" is used in Victorian fiction and to identify recurring textual patterns associated with the fireplace pose. By analysing concordance lines through the KWICGrouper, I aimed to understand how physical positions around the fireplace symbolise authority, domesticity, gender identity, and social hierarchy within nineteenth-century literature.

Understanding the "Fireplace Pose"

The activity is based on the research of Barbara Korte (1997) and Michaela Mahlberg (2013), who observed that many nineteenth-century novels repeatedly depict male characters standing with their backs to the fire, whereas female characters are more often portrayed sitting near the fireplace.

This recurring bodily posture is not merely descriptive but culturally significant. Since Victorian society associated the domestic interior with social order and gender expectations, standing before the fireplace symbolised confidence, authority, and masculine power. Women, constrained by contemporary ideas of propriety and practical concerns such as long dresses catching fire, were generally represented in seated positions.

Thus, the fireplace becomes more than a physical object; it functions as a cultural symbol reflecting Victorian ideologies.

My Exploration Using CLiC

I began by selecting Dickens' Novels (DNov) within the CLiC corpus and searched for the keyword "fire." The concordance returned more than 1,700 occurrences, making it difficult to examine every instance individually.

To narrow the results, I used the KWICGrouper, adjusting the search span from L5 to L1 and entering the word "back." This filtered the concordance and highlighted expressions such as "his back to the fire" and similar constructions.

The KWIC interface allowed me to observe repeated linguistic structures that would have remained unnoticed through ordinary reading.

Observations from Dickens' Novels

While examining the concordance lines, I noticed that many male characters were repeatedly positioned with their backs to the fire during important conversations or moments of authority. These descriptions frequently appeared in domestic interiors, suggesting that the fireplace functioned as the symbolic centre of Victorian household life.

Several concordance lines also connected the fireplace with reflection, conversation, social interaction, and emotional intensity. Rather than serving merely as background description, the fireplace often shaped the atmosphere of important narrative scenes.

The repetition of this posture across different novels indicates that Dickens consistently employed bodily positioning as part of his characterization.

Comparing Dickens with Other Nineteenth-Century Novelists

The second part of the activity involved repeating the search within other nineteenth-century novels available in CLiC.

Although references to fire continued to appear, I observed that Dickens seemed to employ the fireplace pose more consistently than many other Victorian authors. In Dickens' fiction, standing before the fireplace frequently accompanies scenes of authority, discussion, and family interaction, whereas in other novels the descriptions appeared comparatively less formulaic.

This comparison demonstrated how corpus tools can reveal an author's stylistic preferences through quantitative evidence.

My Personal Experience

Before using CLiC, I generally approached literature through traditional close reading, focusing primarily on themes, symbolism, and character analysis. This activity introduced me to an entirely different perspective by showing how computational analysis can uncover hidden textual patterns across thousands of pages.

Initially, the concordance interface appeared technical and overwhelming because of the large number of search results. However, once I learned to use the KWICGrouper, the activity became highly engaging. Filtering the data and observing repeated expressions made me realise that literary language often follows systematic patterns that readers rarely notice during ordinary reading.

What fascinated me most was discovering that something as ordinary as a character's physical posture could reflect Victorian social values concerning gender, authority, and domestic life. The activity demonstrated that literary description is never entirely neutral; even small narrative details contribute to broader cultural meanings.

This experience also strengthened my appreciation for Digital Humanities as an interdisciplinary field that combines literary interpretation with computational analysis.

Critical Reflection

The fireplace pose illustrates how language constructs social identity rather than simply describing physical actions. Corpus analysis makes these patterns visible by examining frequency, repetition, and contextual usage across multiple texts.

The activity also challenged my understanding of literary criticism. Rather than replacing traditional interpretation, computational analysis provides empirical evidence that supports critical arguments. Close reading explains why a passage is meaningful, whereas corpus analysis demonstrates how frequently particular patterns occur throughout an author's work.

Consequently, Digital Humanities encourages a balanced methodology where quantitative analysis complements qualitative interpretation.

Learning Outcomes

This activity significantly enhanced my understanding of Digital Humanities by providing both theoretical knowledge and practical experience in corpus-based literary analysis. Through the use of the CLiC platform, I learned how to perform concordance searches, interpret KWIC (Key Word in Context) displays, and use the KWICGrouper to identify recurring linguistic patterns across literary texts. It deepened my understanding of Corpus Stylistics by demonstrating how repeated textual patterns can reveal important cultural meanings and support literary interpretation. The activity also helped me appreciate the symbolic significance of the Victorian fireplace pose as a reflection of gender roles, authority, and domestic life in nineteenth-century fiction. Most importantly, I realised that computational methods and traditional close reading are complementary rather than opposing approaches to literary criticism. This experience strengthened my confidence in using digital tools for postgraduate research and encouraged me to adopt interdisciplinary methodologies that integrate literature, language, and technology for more comprehensive literary analysis.

Conclusion

This CLiC activity demonstrated that Digital Humanities offers innovative methods for studying literature without diminishing the importance of human interpretation. Through corpus analysis, I discovered that recurring descriptions of characters standing with their backs to the fire are not accidental narrative details but culturally meaningful representations of Victorian gender roles and social hierarchy.

The experience revealed that computational tools such as CLiC enable readers to identify hidden textual patterns that remain invisible during conventional reading. As a postgraduate student of English Literature, I found this activity intellectually rewarding because it expanded my understanding of literary analysis beyond traditional interpretative approaches. It showed that the future of literary scholarship lies not in choosing between close reading and digital methods, but in integrating both to achieve richer and more comprehensive interpretations of literary texts.

Exploring 1984 through Voyant Tools: A Digital Humanities Experience

Text Mining and Visualizing George Orwell's 1984

What is Voyant Tools?

Voyant Tools is a free web-based text analysis platform widely used in Digital Humanities, corpus linguistics, and literary studies. Developed by Stéfan Sinclair and Geoffrey Rockwell, it enables researchers, students, and scholars to analyse literary and non-literary texts using computational methods. Instead of relying solely on traditional close reading, Voyant supports distant reading, allowing users to identify patterns, word frequencies, relationships, and thematic trends across an entire text.

Voyant transforms textual data into interactive visualizations that help users explore vocabulary, recurring themes, character relationships, and linguistic structures. These visual tools provide empirical evidence that complements literary interpretation and enables readers to approach texts from new analytical perspectives. For this activity, I uploaded George Orwell's 1984 and explored the novel using several visualization tools available in Voyant.

Tools Explored in Voyant

During this activity, I used the following visualization tools to analyse 1984:

1. Cirrus

The Cirrus tool generates a word cloud displaying the most frequently occurring words in the text. Larger words represent higher frequency, enabling readers to identify the dominant themes and vocabulary of the novel at a glance. In 1984, words such as Winston, Party, Big Brother, thought, and room emerged prominently, reflecting the novel's central concerns with surveillance, power, and political control.

2. Constellations

The Constellations tool visualises relationships between frequently occurring words by displaying how they are connected within the text. It helps readers observe associations among characters, concepts, and recurring ideas. This interactive network illustrated how keywords like Party, Winston, Julia, truth, and Big Brother are interconnected throughout the novel.

3. Links

The Links tool presents word associations in the form of a network graph, illustrating how different terms frequently appear together. This visualization helped me identify important thematic relationships, particularly those connected with surveillance, authority, resistance, and political ideology. It demonstrated how Orwell repeatedly connects specific concepts to reinforce the oppressive atmosphere of the novel.

4. DreamScape

The DreamScape visualization presents the text in an artistic and exploratory format, allowing readers to observe textual patterns from a different visual perspective. Rather than focusing solely on frequency counts, DreamScape encourages a more imaginative engagement with the language and structure of the text. It highlighted how recurring motifs contribute to the psychological atmosphere of 1984.

5. Loom

The Loom tool compares the distribution and occurrence of selected words across different sections of the novel. It enabled me to observe how particular themes become more or less prominent as the narrative progresses. This temporal perspective helped me understand the structural development of Orwell's novel and the changing emphasis on concepts such as surveillance, rebellion, and truth.

6. Trends

The Trends tool displays graphs showing how frequently selected words appear throughout different parts of the text. By analysing these trends, I observed how important concepts fluctuate as the story develops. This provided insight into the narrative progression and thematic evolution of 1984.

My Experience Using Voyant Tools

Using Voyant Tools to analyse 1984 was an engaging and insightful experience that introduced me to a new dimension of literary analysis. Before this activity, I primarily relied on close reading to interpret literary texts, paying attention to themes, symbolism, and character development. However, Voyant demonstrated how computational text analysis can complement traditional literary criticism by revealing patterns that may remain unnoticed during ordinary reading. The interactive visualizations made the novel more accessible from a quantitative perspective, allowing me to identify recurring words, thematic clusters, and textual relationships with greater clarity.

Among all the visualization tools, I found Cirrus and Trends particularly useful because they immediately highlighted the dominant vocabulary and showed how important themes evolved throughout the narrative. The Constellations and Links tools further enriched my understanding by illustrating relationships between characters and concepts, while DreamScape and Loom offered creative and structural perspectives on the text. Analysing 1984 through these visual representations deepened my appreciation of Orwell's careful use of language and demonstrated how digital tools can provide empirical evidence to support literary interpretation.

Learning Outcomes

This activity significantly enhanced my understanding of Digital Humanities and the role of computational tools in literary research. I gained practical experience in using Voyant Tools to perform text mining, word frequency analysis, and thematic visualization. The activity helped me understand how digital visualizations such as Cirrus, Constellations, Links, DreamScape, Loom, and Trends can reveal recurring patterns, thematic relationships, and structural developments within a literary text. I also realised that computational analysis complements rather than replaces traditional close reading, providing quantitative evidence that strengthens literary interpretation.

Furthermore, this experience broadened my perspective on interdisciplinary literary studies by demonstrating how technology can enrich textual analysis. Working with 1984 enabled me to explore Orwell's themes of surveillance, power, ideology, and language from both qualitative and quantitative perspectives. As a postgraduate student, I gained confidence in using digital tools for literary research and recognised the importance of integrating Digital Humanities methodologies with conventional critical approaches to produce more comprehensive and evidence-based literary analyses.

References

Barad, Dilip. "What if Machines Write Poems." Dilip Barad | Teacher Blog, 21 Mar. 2017, https://blog.dilipbarad.com/2017/03/what-if-machines-write-poems.html. Accessed 3 Aug. 2026.