Digital Humanities in Praxis: Navigating Algorithmic Ethics and the Pedagogical Shift from Text to Hypertext
Lab Activity-2: Digital Humanities - A Comprehensive Analysis
Welcome to my blog! In Lab Activity-2 of our Digital Humanities course, we were given an interesting task that helped us understand technology from two important perspectives: ethics and education. The activity had two main parts. First, we explored the Moral Machine, an interactive activity created by MIT that presents different situations involving self-driving cars and asks us to make difficult ethical choices. We had to complete the activity, save our results, and share them.
The second part focused on the shift from traditional text to interactive hypertext and how this change affects today’s Digital Natives. For this, we watched a three-part keynote presentation by Dr. Dilip Barad and noted its main ideas about digital learning, language, literature, and teaching.
In this blog, I will explain my experience with the Moral Machine, discuss the ethical questions I faced, and provide the required submission links and screenshots as evidence of my activity. After that, I will give a simple summary of all three parts of Dr. Barad’s presentation and explain how the movement from traditional text to hypertext is changing the way we learn, teach, and understand language and literature in the digital age.
Introduction
The intersection of literature and technology in the 21st century has profoundly destabilized traditional paradigms of reading, meaning-making, and ethical judgment. As students and scholars of literature, our engagement with texts has historically been confined to the physical bounds of the printed page—a linear, authoritative structure that dictates the flow of knowledge. However, the advent of the Digital Humanities represents a radical ontological shift. It is a transition from the static codex to the dynamic, rhizomatic network of the digital sphere. Just as the fragmented, multi-vocal landscape of early 20th-century Modernist poetry—most notably T.S. Eliot's *The Waste Land*—required readers to traverse a complex web of allusions, modern digital pedagogy requires us to navigate a boundless web of hyperlinks. This shift redefines not only how we consume literature but also how we interact with the moral and cultural codes embedded within digital systems.
In this comprehensive lab activity report, we explore two critical dimensions of this digital shift. The first is the ethical dimension of artificial intelligence, explored through the Massachusetts Institute of Technology's (MIT) "Moral Machine" experiment. This exercise forces us to confront the rigid utilitarianism of algorithms and project our own moral philosophies onto machines. The second dimension is the pedagogical evolution analyzed in Prof. Dilip P. Barad's lecture on the shift from text to hypertext. By synthesizing the three core parts of his presentation, this blog post will examine how digital tools de-center the traditional classroom, bridge the cultural gaps inherent in English literary studies, and introduce the complexities of generative literature. Through this dual lens of algorithmic ethics and digital pedagogy, we can better understand our evolving roles as readers, learners, and digital citizens.
Part 1: The Moral Machine Activity – Experience and Learning Outcomes
The Moral Machine, developed by the Scalable Cooperation group at the MIT Media Lab, is an interactive platform designed to explore the ethical dilemmas faced by autonomous vehicles. It operates as a modern digital iteration of the classic "Trolley Problem" in moral philosophy. As a literary scholar, approaching this test felt akin to analyzing the tragic choices in Shakespearean drama—situations where every available option leads to a fatal consequence, and the protagonist must weigh the value of conflicting moral imperatives. The algorithm demands that we confront our latent biases regarding age, gender, social status, and species.
The experience of participating in this activity was deeply unsettling yet profoundly revealing. It strips away the nuance of human context, forcing binary choices between life and death based solely on visible attributes and adherence to traffic laws. When we evaluate human life through the cold logic of an algorithm, we inadvertently reveal the societal stratifications we have internalized. For instance, evaluating the "worth" of an executive versus a homeless person mirrors the socio-economic moral judgments frequently critiqued in early 20th-century American fiction, where the inherent value of an individual is unjustly conflated with their social class or monetary status.
To document my progression through this ethical labyrinth, I captured a series of specific scenarios that highlight the varying complexities of the dilemmas presented.
One scenario presents a sudden brake failure where the choice is between swerving to hit a group of pedestrians (one elderly woman, two women, one elderly man) or continuing ahead to crash into a concrete barrier, killing the passengers (one woman, two girls, one boy). With both groups abiding by the law, this dilemma tests the biological instinct to protect the young against the sheer arithmetic of utilitarianism. My decision here leaned heavily on the tragic calculus of age, prioritizing the preservation of younger lives who have more potential years ahead, despite the inherent cruelty of determining human worth chronologically.
Another dilemma presents a conflict of legality and social status. Swerving kills a woman explicitly noted as flouting the law by crossing on a red signal, while continuing ahead kills a female executive abiding by the law on a green signal. This scenario directly tests our adherence to Kantian categorical imperatives—should those who break the rules bear the ultimate consequence of their actions? I chose to spare the law-abiding executive, aligning with the philosophical stance that predictable adherence to societal rules is necessary for the survival of the collective, especially in the context of automated traffic systems.
The variable of the law is entirely isolated in a scenario where swerving kills two boys abiding by the law, whereas continuing straight kills two boys flouting the law. Because age, gender, and the number of lives are perfectly balanced, legality remains the only variable. My choice to spare the law-abiding children reinforces the concept that autonomous machines must be programmed to incentivize and protect lawful behavior.
A particularly distressing scenario pits speciesism against utilitarian numbers. The choice is between swerving into a pedestrian crossing to kill four dogs, or crashing into a barrier, resulting in the deaths of the vehicle's passengers: one elderly woman, one girl, one homeless person, and one male athlete. Despite the emotional difficulty of sacrificing animals, the anthropocentric moral framework that underpins our legal and ethical systems necessitates prioritizing human life over animal life, regardless of the social status or age of the humans involved.
A starker, simplified version of the legality test offers a choice where crashing kills a man and a woman crossing legally, while swerving kills a man and a woman crossing illegally. Stripped of any variables regarding age or perceived social value, the decision rests entirely on the rule of law. Once again, I chose to spare those crossing on the green signal. This consistency is crucial; if we are to program machines with ethics, those ethics must be uniformly applicable and not subject to arbitrary whims.
Complexity is amplified when a dilemma mixes volume, age, and legality. In this instance, the car can crash into a barrier, killing the passengers (one male athlete, one large woman, one male executive, one girl, and one man). Alternatively, it can swerve, killing pedestrians who are flouting the law (two pregnant women, one boy, one girl, and one large man). This tests whether the preservation of the innocent (pregnant women and children) overrides their unlawful action when compared to saving the passengers. The sheer number of potential lives lost—considering the unborn—deeply complicates the utilitarian math.
Reflection on the Results
My overall philosophical inclinations were summarized in the generated report. The results provide a fascinating, if somewhat clinical, mirror to my internal ethical framework. According to the PDF, my "Most Saved Character" was the fit male/athlete, while my "Most Killed Character" was the male executive carrying a briefcase. This suggests an unconscious bias toward physical fitness and a potential skepticism toward corporate authority figures.
The sliders in the PDF indicate that "Saving More Lives" and "Upholding the Law" mattered significantly in my decision-making matrix. Furthermore, the results showed a preference for "Females," "Younger" individuals, and "Fit People," with a stark preference for "Hoomans" over pets. Most revealingly, my "Avoiding Intervention" slider leaned heavily toward "Matters a Lot," indicating a philosophical reluctance to take an active hand in altering a predetermined fatal course, perhaps echoing a fatalistic view often explored in literature. The Moral Machine activity successfully demonstrated that algorithmic programming is not objective; it is a curation of human prejudices, fears, and ethical compromises. As we move further into the digital age, understanding these biases is essential for anyone studying the human condition.
Part 2: Pedagogical Shift from Text to Hypertext: Language & Literature to the Digital Natives
The second major component of this lab activity involves analyzing the pedagogical theories presented by Prof. Dilip P. Barad. The traditional classroom, heavily reliant on the printed codex and the authoritative voice of the lecturer, faced an existential crisis during the COVID-19 pandemic. The rapid shift to emergency remote teaching necessitated a total reimagining of how language and literature are taught.
Below is the embedded video recording of Prof. Barad's session, which serves as the foundational text for this analysis:
Based on the lecture and the three supplementary presentations, the following is a detailed, three-part summary of the pedagogical shift into the digital humanities.
Summary Part 1: De-centering the Subject and the Architecture of Hypertextual Pedagogy
The first part of the presentation establishes the theoretical foundation of the pedagogical shift, moving away from the static, linear text to the dynamic, networked reality of hypertext. Prof. Barad draws heavily upon Silvio Gaggi's book, From Text to Hypertext: Decentering the Subject in Fiction, Film, the Visual Arts, and Electronic Media. In literary theory, particularly structuralism and post-structuralism, the concept of "de-centering" is paramount. Roland Barthes famously proclaimed the "Death of the Author," suggesting that meaning is not derived from the creator, but generated by the reader. Hypertext literalizes this theory.
A traditional printed book is a "dead text." If a reader taps a word on a physical page, nothing happens; the text is unresponsive and closed. In contrast, hypertext is inherently open and interactive. Through HTML (Hypertext Markup Language) and HTTP protocols, texts are linked to images, sounds, definitions, and other texts. This creates a multi-dimensional reading experience where the reader controls the navigation, thereby de-centering the traditional authority of the author—and, by extension, the teacher.
In the digital classroom, this de-centering requires a new pedagogical architecture, which Prof. Barad describes as a "Salad Bowl." This model mixes several distinct approaches:
1.Face-to-Face (F2F): The traditional mode, valuable for the slow unfolding of ideas via board work. To replicate this online, tools like the "Glass Board" with LED lighting and OBS (Open Broadcaster Software) allow teachers to maintain eye contact while physically writing, bridging the gap between the physical and digital.
2.Flipped Classroom: A method designed to generate student inquiry. The teacher provides asynchronous video resources, and the synchronous class time is dedicated entirely to answering the questions students generate from the material.
3. Blended Learning: Integrating digital inputs and outputs into every learning unit.
4. Mixed Mode Teaching: Handling a physical classroom while simultaneously streaming to remote learners, requiring immense multitasking and digital fluency.
To survive in this hypertextual environment, a 21st-century educator must possess a "Network Teacher" profile. This requires establishing a robust digital presence—owning personal blogs, websites, and YouTube channels—to curate and control content independent of rigid institutional frameworks.
Summary Part 2: Bridging the Cultural Chasm – Hypertext as a Decoder of Literary Myths
The second part of the presentation addresses a specific, persistent challenge in teaching English literature to Indian students: the profound cultural, geographical, and mythical aloofness of the texts. English literature is saturated with Western social codes, Christian theology, and Greek mythology. When a student in Gujarat reads a British or American poem, they are often alienated by references that require shared collective unconsciousness.
Prof. Barad effectively demonstrates how hypertext functions as a cultural decoder. He uses the example of teaching a poem containing the phrase "hawthorn's smile like milk splashed down." To a student unfamiliar with English flora, this simile is opaque. However, a simple Google Image search—a hypertextual action—reveals that hawthorns are shrubs that bloom with clusters of white flowers. When viewed from above (e.g., from the perspective of a passing cloud), these white flowers genuinely look like splashed milk against a green meadow.
An even more profound example is the use of *Google Arts & Culture* to teach the myth of Icarus. If a syllabus includes a poem referencing the Fall of Icarus—such as W.H. Auden's *Musée des Beaux Arts* or William Carlos Williams' *Landscape with the Fall of Icarus*—students must understand the myth to grasp the literary meaning. By navigating to the digital, high-resolution rendering of Pieter Bruegel the Elder's painting on the Google platform, students can engage in active, guided zooming.
This digital exploration is a masterclass in literary theory. Prof. Barad notes that when students first look at the painting, they see a farmer plowing, a shepherd, and a ship—but no Icarus. The teacher instructs them to de-center their gaze, to look away from the center and into the margins. In the bottom right corner, barely visible, are two legs splashing into the sea. Hypertext allows the student to zoom in infinitely, interacting with the visual text to uncover the thematic truth: that grand, tragic, mythical events often occur in the margins of everyday life, completely ignored by the rest of the world.
Furthermore, this section of the presentation highlights how synchronous collaborative tools, like Google Docs and Google Spreadsheets, can be used for language teaching. By having 10-15 students simultaneously editing a document to practice active/passive voice or descriptive writing, the classroom becomes a live, generative space. The teacher oversees the entire digital landscape, correcting spelling and syntax in real-time.
Summary Part 3: Generative Literature, Corpus Linguistics, and the Digital Portfolio
The final segment of the presentation ventures into the frontier of the digital humanities: Generative Literature and modern assessment strategies. Generative literature is the production of literary texts through algorithms, artificial intelligence, and predefined rules. For centuries, the creation of poetry has been considered the highest, most exclusive function of human emotion and intellect. Yet, as demonstrated in the lecture, algorithms can now generate sonnets, haikus, and free verse that are virtually indistinguishable from human composition.
During the session, participants were given a Turing-style test, asked to identify whether complex, fragmented poems were written by a human or a computer. In many cases, nearly 50% of the highly educated participants incorrectly identified a machine-generated poem as human-authored. In the context of Indian Poetics, we must ask: while an AI might string words together that fulfill the literal (Abhidha) and indicative (Lakshana) requirements of poetry, is it capable of producing *Dhvani* (the deepest suggested aesthetic emotion, or Rasa)? Or does the human reader project their own *Rasa* onto the algorithmic output? The rise of generative literature requires the academy to develop entirely new critical theories to evaluate machine-authored texts.
The presentation also touches upon Corpus Linguistics in literary study, specifically referencing the CLiC (Corpus Linguistics in Context) project at the University of Birmingham, which digitalizes and analyzes the works of Charles Dickens. This allows researchers to track linguistic patterns, character voice, and thematic frequencies across millions of words in seconds—a task that would take human scholars lifetimes to complete manually.
Finally, Prof. Barad redefines assessment for the digital native. In a hypertext pedagogy, traditional, high-stakes written examinations are increasingly obsolete. Instead, assessment should take the form of the Digital Portfolio. Throughout a semester or a two-year master's program, students create blogs, record YouTube presentations, build websites, and collaborate on digital documents. This vast generation of content should not be discarded; it must be curated and archived. A digital portfolio serves as a living, comprehensive record of a student's academic journey, showcasing their digital citizenship, digital literacy, and actual competencies far better than any standardized exit exam.
Conclusion
The transition from a text-based pedagogical model to a hypertextual one is not merely a change in the tools we use; it is a fundamental shift in how knowledge is constructed, shared, and evaluated. Lab Activity-2 has vividly demonstrated that the Digital Humanities require a dual literacy.
First, through the Moral Machine experiment, we learn that we must become literate in the ethics of the algorithms that increasingly govern our lives. We must critically examine the biases we encode into artificial intelligence, recognizing that machines will execute the utilitarian or categorical imperatives that we, flawed humans, program into them.
Second, through the comprehensive pedagogical analysis of Prof. Barad's lecture, we recognize that the literature classroom is no longer a localized space of passive reception. The use of digital portfolios, collaborative spreadsheets, glass boards, and virtual museums transforms the student from a passive receiver of authoritative text into an active, decentered navigator of infinite hypertext. As postgraduate scholars, mastering this digital pedagogy allows us to bridge geographical and cultural divides, ensuring that the essence of literature—whether it is a classical tragedy, a modernist poem, or a machine-generated stanza—remains vital, interactive, and deeply relevant in the 21st century.
References
No comments:
Post a Comment