Digital History

Mapping the Past and Historical GIS

When I began this exercise, I wanted to map the first few individuals that had positive COVID-19 cases in Toronto. However, I quickly faced some challenges during my research due to unreliable or lack of informative sources. Many of these sources were inconsistent and I felt that relying on them would have resulted in a sparse, and potentially inaccurate map.

This topic is a sensitive one at that, and I wanted to make it as accurate as possible. The only repetitive case documented in any source begins with a man in his 50s, who travelled back with his wife from Wuhan, China in January 2020. With this data as a starting point on a map, told very little to tell a broader story of a deadly pandemic we all experienced.

This led me to increase the geographic scale to the whole province of Ontario.

After looking at numerous numbers of different sources, I finally found a promising one: a dataset published by the Government of Ontario that tracked nearly a million cases from 2020-2024. I aimed for a story mapping focus of administrative geography. I think mapping adds dimension that the dataset contents could not reveal. Although the one I created is an amateur's work, I hoped to connect the PHUs who reported their cases to show the virus' travelling. I also found the region who had the highest number coronavirus reports: Toronto Public Health. This would make sense for a densely populated city and everyone being in closer proximity.

I think spatial history emphasizes connection, showing what was happening in all places in a range of time. Additionally, I think that digital tools pose the challenge of choosing what data is left out for the visual narrative because of the geographic focus.

I know this was simply an exercise but if I were to do a full project on this topic, I think I could create a great story map with just using ArcGIS.

Analyzing a historical podcast

Consuming History in the Digital Age

When I was young, I learned history through shows such as the History Channel and films that re-tell history in a lens digestible for children. These shows had an authoritative tone, seemed refined and thus felt trustworthy, embodying the broadcast era.

In modern times, new digital platforms such as social media (TikTok, YouTube, Instagram, Podcasts and Documentaries on Disney Channel or Netflix, Ted Talk) have expanded the range of access to indulge in different content, educational or just simple cooking recipes or gossip.

In my daily life, I often encounter historical information or true crime cases while doom scrolling. In this genuine experience, storytelling remains at the heart of how we can connect with the past. Podcasts I preserve the tone or narrative style reminiscent of a grandmother's stories. The emotional engagement in historical podcast matters enormously because of the inherent sensitivity in historical topics.

Cooking dinner, listerning to murder

I am a big fan of listening to Ted Talks and crime podcasts while doing mundane activities like cooking. The history it speaks about emotionally resonate with me on a personal level. But I also find that the speed and algorithmic influence shape how and what I see online. These platforms offer quick consumption, a 60-second summary of an event that cannot capture nuance. These algorithms maximise not accuracy but trapping the viewers in a filter bubble for engagement. There is a high potential for misinformation that lurks around.

Educators must prepare their students to carefully and responsibly navigate this domain. If we teach critical habits such as questioning what we see online and what it's purpose might be, we can then analyze how these narratives can manipulate emotions. Audiences should practice lateral reading for the goal of transforming how we consume history online.

A Digital History Review

Rotten Mango Podcast, Episode 328: "The Killer Who Vlogged His Murder Plan For 8 Years"

Is this history? A Necessary Question

Before I begin my review, I will address the fundamental question: Does this episode, specifically, present historical content?

I argue that yes, it does. The Rotten Mango Episode 328: "The Killer Who Vlogged His Murder Plan For 8 Years", deals with a past event. It is the Peter Keller case, that unfolded in 2012. The source itself consults a variety of sources (news articles, the Keller video diaries, court records) and engages with a broader historical context (survivalist culture of early 2010s and the 2012 doomsday craze). It draws on both primary and secondary sources, provides citation for any evidence and claims, and most importantly, offers an interpretation of why such an event unfolded, illustrating historical practices.

As a consumer, I have creative liberty in my engagement. The production team presents the historical information in an entertaining way. Stephanie Soo's crime podcasts contain quality history, and I think this episode and many others, makes for a perfect artifact in analyzing how millions of people encounter history today.

The Episode

This episode presents the 2012 case of Peter Keller, a Washington survivalist who murdered his wife Lynette and daughter Kaylene before fleeing into a bunker, he spent 8 years carving in the mountains. Soo argues that he was a raging narcissist who viewed his family, simply as an extension of himself, noting that his video diaries capture and reveal Keller's obsession with himself.

The information is communicated to the audience clearly, in an entertaining tone. I find that Soo breaks down the complex elements very well, such as distinguishing Doomsday "preppers" and "survivalists" (noted around 30 mins into the podcast).

Soo and her husband take on a conversational tone in almost all her episodes and I think strengthens intimacy. It's as if I am listening to a friend telling me a story. I also find it nice when her husband, asks questions, which makes it more of a casual conversation. However, this format informing about a sensitive topic creates a jarring feeling and the sponsor reads in the podcast formulate uncomfortable juxtapositions. I will argue that the narrative momentum tends to oversimplify the complex reality.

Soo's episode demonstrates transparency, and a strong one at that. She will let the audience know if she was not able to verify an information. Her production team also note this on the episode note, modeling responsible practice.

I really enjoy listening to her episodes because of the suspense, the audio use and emotional appeal. She invites me to do a deeper investigation on my own.

Many of her episodes represent a memorial for the victims of the crimes and she collaborates with organizations to establish scholarships. In this way, I think digital history has power and peril in how accessible it is, while maintaining transparency.

For millions of people, including myself, as cooking dinner while listening to murder, her stories function as history. She performs the role of an educator that equips her audiences with tools to consume thoughtfully. Every true crime story had real victims whose lives mattered.

When I queue up Rotten Mango while cooking, it's like participating in a modern practice of historical consumption. I am also, whether I realize it or not, engage in an act of critical awareness. The intimate storytelling and emotional engagement carry this weight of algorithmic influence, speed and potential misinformation.

The episode presented demonstrate the power and responsibility of digital history, and as Stephanie Soo and her production team craft each new topic, they never forget to humanize victims and cite sources with full transparency. Yet the episode's entertainment format will occasionally oversimplify complicated truths. But this is not a condemnation of the podcast, rather a reality of how many of us currently encounter the past.

Using AI to “interpret” the past / Responsible use of generative AI

Why Historians Aren't Going Anywhere

I remember watching The Imitation Game, the film starring Benedict Cumberbatch as Turing. The film stuck with me, not just for the test but for showing Turing's own humanity that went unrecognized by the state, who also persecuted him for being gay. It doesn't make sense to me because he designed a machine that could imitate conversation, yet society couldn't see him.

When Dr. Humphries' lecture mentioned the Turing test, my mind kept coming back to the irony: if a machine can convince you it is human, should we consider it intelligent?

The Turing story asks a deeper question: whose humanity do we choose to see, and whose do we choose to erase?

From the class demo of testing a chosen AI with historical prompts, I navigate and reflect on this question. I used Claude AI (Sonnet 4.6, the free version) and saw for myself what AI can really do on par to historians, and what it missed.

What Learned from Claude AI

(What AI Can Do (And Does Well)

The topic of the French Revolution in and of itself was a daunting one but asking Claude to summarise the topic, produced a response that was chronological and neutral. Claude walked me through different schools of thought and named key historians, and incorporated a note that no single-cause model holds up today.

This was incredibly useful for someone in need of a quick orientation on an overwhelming topic.

I prompted about primary sources as a follow up and Claude delivered with a breakdown of the different forms. It provided options from cahiers de doléances (notebook for ordinary people's grievances), provincial archives (for studying counter-revolution and police reports (that tracked what ordinary people thought). I think this was helpful because it specified what would be the most "helpful" in interpreting voices from both bourgeoisie and proletariat lens.

I created a new chat, I uploaded texts by Lucretia Mott and Elizabeth Cady Stanton. I prompted Claude to find recurring themes, and for each of them, attaching a quote from the corpus. Claude provided me with seven: Innate Equality, Denial of Rights, and others.

In prompting about gender-history perspective, Claude produced an analysis that introduced new concepts I did not know about such as "coverture" (a legal doctrine where a woman's rights were absorbed by their husband) and the "separate spheres" ideology.

Claude AI was powerful in summarizing debates, pointed me towards multiple sources it found "relevant" or "useful". It spotted patterns across documents and applied academic frameworks.

But becuase I provided clear instructions, with specific context and also prompting it to take on a persona, it was able to give me a sophisticated output.

So, what is AI doing? AI is working with stuff that already exists summarising, retrieving, spotting, and applying. This is not new in research, it doesn't offer new ideas, genuine insights, nor is it surprised by evidence. It takes no stake in anything. It's like an RA that's read everything but never felt anything.

Answering the Three Questions

What kind of tool is AI?

Based on the demo, AI is an RA and a tool for finding patterns, not a co-author. When I pushed for an original interpretation, it still only gave me synthesized patterns. AI's power stops at execution, but all responsibility remains entirely ours.

What should we not ask AI to do for us?

To never outsource our core original research because once it is uploaded, your data trains future models, often without informed consent.

That work becomes a data point for the big tech companies. Additionally, never ask AI to genuinely understand perspective we have trained it to see as wrong. For example, our demo contained the prompt to analyze texts from a perspective skeptical of women's suffrage. Claude struggled with this response. It wants to perform the modern values it had been trained on. I think AI can identify the tension but cannot wrestle with it.

Who benefits from it?

The Big Tech Companies control the tools, training, data and pricing. So, each second, we use AI, it's helping to train the next version, almost like free labour to gain access.

The knowledge sourced by AI is from data that's widely available on the internet. This means it overlooks local, and non-digitized histories. This is confirmed by a commenter in Dr. Humphries article. Still, generative AI's convenience gives a huge amount of power to the big tech companies that own LLM.

Is this really the last generation of historians?

An analysis of Gans and Humphries

I think the Turing test forces us to critically assess if AI becomes indistinguishable from a human, will society still value human "detective work" (mentioned by Dr. Humphries in the comments)? Or is the output all that matters, thus making the human historian economically obsolete, such as Dr. Humphries' woodworking hobby competing with IKEA?

Both authors spent most of their post demonstrating how powerful AI has become. 

Joshua Gans demonstrated that it was able to publish a peer-reviewed paper with the help of AI. Humphries watched as AI navigated across 21 archives on its own. So, the evidence is right there, AI can do research. 

But I think in both of these posts, the things that the authors felt they had to add, almost as an afterthought, were relevant.

For Gans it was the disclaimer about taking full responsibility because in including that disclaimer, he and the journal knew that the accountability cannot be outsourced even if AI did the work. If someone questions the findings, they won't email OpenAI, they'll email Gans. 

For Humphries he describes Deep Research hallucinating, fabricating quotes and even producing codes that don't work. His conclusion reminds us to treat LLM's critically.

Critical engagement...that's a historian's job. 

Even when AI is the most impressive, it still needs to have humans catch its mistakes, take responsibility for its outputs and decide what questions matter in the first place. 

The future of historical research looks like a partnership to me. 

If we let AI do the thinking and taking the responsibility, we don't just lose our jobs, we also lose accountability, we lose human willingness to revise, to sit with the discomfort and to be wrong. 

History becomes a product without a producer. 

So, to answer the question: is the last generation of historians...?

That depends on whether we noticed what Gans and Humphries themselves almost seem to hide in plain sight, and that's the failures and the moments where AI broke and the human had to step in. They're not just footnotes, they're the whole point. 

A Personal Response

The imitation of a true human craft cannot be the same as understanding it. Whether it is a table or its human conversations, if we value only the output, we lose what is irreplacable.

I think AI does not need to match the "best" historian, it just needs to be "good enough" for everyday tasks as IKEA can simply be "functional" for most people.

If it can do that, then the demand for historians could shrink.

I return to Turing. That story reminds us that history is more than facts. It is about identity that shapes lives and legacy. AI can summarize it in seconds, but the algorithm cannot comprehend it. The test can measure imitation, but history requires empathy, lived experience and a recognition of our own humanity. For that, we need human historians. 

Final Thoughts

After running Claude through historical prompts and reading Gans and Humphries, I land on the fact that AI is powerful. It's able to summarize debates I know nothing about and point me towards archives I didn't even know existed, and spot patterns across texts I would probably miss on my own. It's a legitimately useful tool for research. 

However it cannot take responsibility. It cannot wrestle with perspectives it's been trained to see is wrong and it cannot decide whose stories matter. 

The Turing test measures imitation. Can a machine convince you that it's human? By that metric, AI is close.

Working with Voyant Tools to Create Data Visualizations

For this exercise, I experimented with Voyant Tools and used Dostoevsky's novella Notes from Underground, sourced from Project Gutenberg.

I've actually read this novella in my first year Philosophy class, I read it again from time to time.

I chose this text because Fyodor Dostoevsky's novella is unique case study. The protagonist, the infamous "Underground Man" is extremely self-aware, but trapped and constantly contradicts himself. So, if there is any literary work that could benefit from having the linguistic pattern quantified and visualized, I thought this would be a good one.

With a document that contains a total of 44,481 total words and 7,479 unique word forms, the corpus was substantial enough to see patterns, but still remain meaningful for my first distant reading method.

First Impressions: The Interface

A feature I found to be distinct in Voyant Tools, is its multi-panel dashboard. Each corner of the site displays a different tool I could reorganize and customize. I actually thought it was just a visual quirk. But now I find it as a philosophical approach to text analysis. The multiple perspectives need to be visible simultaneously, which can help with comparative thinking.

Each panel can be swapped and interactive, which allows me to customize my analytical workspace. Akin to a digital historian's workstation.

The design of the interface is a methodological lesson, and my analysis worked best when I could see the different types of evidence side by side.

Source Criticism

Before any analysis could start, I had to clean the data. The Summary Tool displayed an immediate issue with my corpus and that is the word "gutenburg", which appeared 75 times. This was not in Dostoevsky's lexicon, but a metadata embedded by Project Gutenberg.

My distant reading began with criticising my source by filtering out administrative words. The Stopwords Tool is in charge of this, which affects the word cloud and the frequency statistics in real time. This digital cleaning tool is equivalent to removing stamps or other residues from a physical source or document.

All sources, digital or not, often come with baggage.

Cirrus and Document Terms

After cleaning the data, I analyzed Cirrus (word cloud) and the Document Terms tool to identify which words appeared frequently.

"Man" used 107 times.

"know" used 88 times.

"time used 75 times.

"like" used 75 times.

"love" used 67 times.

I think this highlights the narrators preoccupations: his identity, epistemology, temporality and affect.

The world cloud allowed me to take this data and turn it into a visual map. With the narrator using "man" 107 times, it meant more than just a word. This repetition shows me that questions of identity, self, and what it means to be human are central in the novella. I could narrow my focus for close reading, and philosophical or historical questions.

The Context Tool was like a highlighter, finding every single time that the word "man" was used and displaying sentences it appeared in. For instance, "I am a sick man", found in Part I.

This data no longer showed me a number but a pattern where the narrator used it to describe himself negatively. My analysis concluded that the Underground Man's negative labels for himself might reveal notions of masculinity and identity in 19th-century Russia. The Context Tool didn't directly answer this for me but now I have textual evidence. I could focus on interpretation.

Fancy Visual Voyant Tools: Knots and Dreamscape

After trying the basic tools, I jumped into the fancier tools Voyant offered, such as Knots and DreamScape. These were the most appealing to me.

Knot is a tool that turns the words into spiral or twisty lines. The frequency of the word's usage, the more twists and turns the line will display. The word "know" used 88 times, had more tangles, compared to "man", used 107 times. It's a weird but also cool way of seeing the repetition without the use of numbers.

Discovering the DreamScape Tool was the most exciting for me. It's aim is to find a location's name within the corpus text and dot it in a map. I originally thought it would show me Russia, since the context of the novella takes place there. Instead, it showed multiple locations in North America, places that were not in the book at all.

Before I experimented with the tool it did give me a warning prompt about the tool undergoing experimentation and "do not trust this data" on the Voyant Tools Help page. This taught me to always double-check what the digital tools show me.

Just because a digital tool looks cool, it doesn't speak for its reliability. If Voyant did not show a warning, I would have plainly believed the tool's capabilities of factual data.

For a historian, one shouldn't plainly believe any information at face value. We are surrounded with forged or fabricated documents. It exists in the digital world too.

Final Thoughts

I think Voyant is a helpful pair of glasses for near-sightedness, but not a magnifying glass. If I take the glasses off, I can read the small prints myself, analyse the tone and feel what the underground man is trying to tell me.

I think Voyant is a tool that can be used at the beginning of any project that could help me decide where my focus should be. But the "thinking" job is for me to do on my own with the data it can provide for me.

Activity 1

Website Design / Digitization, OCR, and Text as Data

I worked with two very distinct kinds of digital sources for the first exercise. I never looked at these kinds of sources deeply until now, or moreso, I knew within me the differences, but comparing them was something I would not have done on my own. As someone born in the growing digital age, I'm used to having these sources accessible to me.

Despite both being accessible online, they perform different functions, are created, preserved, and experienced in their own ways.

Digital history is mediated, and I am made more aware of this fact, as I explore the two sources. I find that they are both authoritative in functions. They are both searchable, organized, and widely accessible. But because both sources are shaped by the technological assumptions, they answer to decisions made by institutions: the newspaper editors, by the library of congress. It proposes a limit in preservation and choice of what data is accessible for us to view

The first link is a digitalized newspaper from the Library of Congress.

The second, a born-digital website archived by the Wayback Machine.

Both can be freely accessed online.

THE DAILY WORKER (September 1, 1927).

This digital source can be accessed through the Library of Congress's Chronicling America collection. The name itself is interesting. It is to chronicle what was. This specific daily newspaper is political in nature. In a closer read, it focuses on labor movements during that time, depicting the "left" side of politics critiquing capitalism. Even the ads and editorials reflect this ideology.

I find the detailed metadata of Chronicling America to be its strength. I was able to search by date, state, publication, and subject or keyword, making it easy to navigate and locate what I might need.

I was thinking of two things as I explore the site: what is the scope of the archived newspapers in the database? I encoutered only two results with my selected search. Similar to a physical archive, I wonder who made the decision to digitalize these primary sources.

I did a bit of external research about The Daily Worker, and come to find its roots aligning in Communist Party ideology. This version has shaped institutions like the University of Illinois at Urbana-Champaign Library and the Library of Congress. The decision to release these specific newspapers must be catered to diffirent individuals or is it to inform us as well?

In my experience with digitalized newspapers, some can be hard to read due to the scanning quality, or OCR proccessing. But the metadata organization of the site is extremly well done in my opinion. I could read each word, zoom in or out and easily access the provided images.

If this site were to disappear, the contextual information provided will be lost along with it. I must have been ignorant until this point because I never knew this to be the case in the Americas back then. But now I am a little bit informed.

Tumblr, 2007-2013, Internet Archives Wayback Machine

I decided to pick this site because a few days ago I accidentally deleted my Tumblr account containing blog posts from when I was 18 or maybe even younger. I was going through reddit and got informed that this cannot be undone. That account of mine is lost forever from their database. A reddit user recommended the Wayback Machine to recover older posts, but I am unlucky and could not find anything regarding my account. It made me a bit sad because they contained my memories and my younger self blogging and recording.

The born-digital source of Tumblr was created when I was just 6 years old. Since then, it has transformed into a web of sophisticated digital platform capable of using bots. I used the Wayback Machine to explore the archived versions of Tumblr from 2007, 2008, and 2013, 2014, and adding my experience, till now, 2026.

The left side images are from its founding years, and the right ones are from mid-years. This showcases the rapid growth of digital platforms. They evolve based on feedback and user needs. The early version is text-based and minimal, containing what essential tools you need for a blog site.

In just 2008, it becomes sligthly more refined, but simpler compared to 2013-2014. Now, tumblr is one of my favourite blog sites. It is capable to adding embeded links or even inter-linking other sites.

As you know, this blog is done on Tumblr.. :)

The Wayback Machine can only capture a fragment of what the experience might have been with Tumblr, for many individuals. Tumblr has become so much more interactive, with newer design elements, such as replies, reposts, notes. The Wayback seems to have archived up to the year 2016. This explains why I could not find my lost account in the archive. This limited scope of archival data makes me wonder if they would require informed consent to add more about the remaining years. I personally do not remember signing anything about this. Nevertheless, I am indifferent about this decision to share my personal data.

I discovered the structures that shape how I use Tumblr today but not the lived experience it had, though one can simply imagine. The access to the archive, captured only what the Wayback Machine could provide me. It has succesfully captured the preservation of Tumblr's fragments but not its full ecosystem.

Because it is a born-digital archive, Tumblr's content is often personal, or user-generated. This can cause an ethical problem about who's data is preserved and if there are policies or consent involved before it is shared with the internet. Maybe I will check again in tha few years, if they decide to showcase or capture my old blog site on the Wayback Machine.

Activity 2

Reading Like a Machine: OCR in Action

While doing the OCR activity, it emphasized how technology shapes historical interpretation.

When I scanned the left photo, an excerpt of Georgias, it assumed a very clean printed text, since all the lettering and text are the same font and layout. The main text was decoded very easily and when copy-and-pasting, most text is searchable. The smaller fonts in the side margin were not easily captured.

On the other hand, when I scaned my notes from a class, it failed to recognize any of my handwritten notes. The text can be selected but it is not searchable. It produced an unreable text. If I simply look at it, it is legible but the OCR failed entirely.

For many historians, this error matters. If I were to research a source that is handwritten, the OCR may fail to recognize it and it becomes invisible. Valuable information is lost: ideas, names and content. This doesn't mean they did not exist in the past but they are misinterpreted.