Napoleon’s defeat and Lewis’ continuing fight

Inspired by a 150-year-old work of dataviz, I illustrated three Selma — Montgomery marches to memorialise a giant in the American civil rights movement

Yaning Wu
6 min readJul 18, 2021

Please note: this work describes police brutality.

This week, I discovered the genius of one retired French civil engineer who combined geography, military history, and data to document his countryman’s 19th-century pummelling at the hands of the Russian army (and winter).

Charles Joseph Minard, known only by his last name in dataviz circles, created in 1869 what Yale statistician and information designer Edward Tufte called “the best statistical graphic ever produced.”

An image of Charles Minard’s famous “Napoleon visualisation”.

Instead of describing the in’s and out’s of this complex chart, I’ll invite you to read this article, which analyses the work in stunning detail. I was particularly inspired by Minard’s use of multiple axes: horizontal to demonstrate the French army’s advance and retreat, vertical to show the number of soldiers left after each battle, and an additional bottom axis to illustrate the decreasing winter temperatures that led to such loss of life. Given that this was produced during an era of scarce visualisation tools, the cleanliness and clarity of the final product (excepting the extravagant cursive) is even more impressive.

This weekend also marks one year since Atlanta congressman and civil rights activist John Lewis’ death. Born in 1940 to Alabama sharecroppers, he participated in sit-ins and became a Freedom Rider in Nashville while studying there in the 1960s, later working to found the Student Nonviolent Coordinating Committee (SNCC) and forming strong ties with Martin Luther King Jr and his associated Southern Christian Leadership Conference (SCLC). After leaders in the movement praised his dedication, he spoke at the 1963 March on Washington, directly criticising the U.S. government’s inaction amidst Black voter disenfranchisement in a controversial call to revolution.

At midday on March 7, 1965, Lewis joined protest organiser Hosea Williams and churchgoers in Selma, Alabama to march with 600 African-Americans to Montgomery, the state’s capital. They intended to commemorate the killing of activist Jimmie Lee Jackson, who was shot in the stomach by state troopers as he came to the aid of his mother and grandmother during protests the month before, and push for progress on voters’ rights.

This is where my illustration begins.

Image by author.

Less than a mile from the nonviolent protest’s starting place, the marchers were met with tear gas, smoke, and officers on horseback who charged at and beat them.

“Bloody Sunday” ended with the hospitalisation of several protesters, including Lewis, who suffered a fractured skull. They were given little time to flee.

Images of the injured protesters angered the nation, who saw the entire massacre broadcast on television. Two days later, marchers flowed into the city, having been called there by Martin Luther King Jr previously; wanting to avoid their disappointment, he led them to the far side of the Edmund Pettus Bridge once again, where he was also met with police sirens. At this time, he refrained from taking further action.

Image by author.

Finally, after detailed planning and hard negotiations with United States District Judge Frank Johnson, King was granted permission on March 21st to lead more than 3,000 peaceful marchers from Selma to the Alabama State Capitol in Montgomery. He took Lewis and Williams’ intended journey, picking up tens of thousands of supporters along the way.

By September 1965, sitting president Lyndon B. Johnson had signed into law the Voting Rights Act, prohibiting (on paper) racial discrimination in the voting process. This marked a landmark success in the civil rights movement and removed some of the more egregious processes I had learned about in high school such as literacy tests, assessments to prove citizens’ education level so that they could cast their ballots (their questions, besides being difficult to understand even for highly-educated individuals, often had subjective answers and were marked at the discretion of the authority figure reviewing them).

Why Minard?

Given the disparate nature of the groups documented by the visualisations I’ve presented today, you may wonder why I’ve chosen this centuries-old unconventional method of plotting aesthetics to data.

Selma’s protesters were, of course, nonviolent — they were rendered defenseless in the face of their opposition, while Napoleon’s charges were initially well-equipped for their mission. However, both groups advanced and retreated; during “Bloody Sunday”, police pushed protesters back to where they had come from, and King elected to turn around when it appeared dangerous to continue. The movement that he and Lewis spearheaded, though, made great advances when it magnetised thousands on a fifty-mile march, growing instead of shrinking like the French imperial army.

There is a more technical reasoning, too, behind my fascination with Minard here. It is often difficult to convey geographic and quantitative information at once in dataviz because we risk clutter and all sorts of visual faux-pas. This type of chart, however, doesn’t need an underlying map to demonstrate each event’s location in physical space; place names can be labelled just as they would be in a purely cartographic visualisation. The distance between Selma and Montgomery can be conveyed clearly (albeit better than how I have presented it here) using Minard’s example.

Maybe in a few decades, or sometime soon, we can document the protests of this past year like this, too. I hope there are more advances than retreats, and that the retreats are never engendered by such brutality. More than anything, I hope that we never lose sight of both the lengths we have come and the greater distance still left to travel.

Data sources:

These are hyperlinked above, but I found them insightful as a newbie to (or rather, someone who has forgotten most of) American history.

I would also recommend accessing primary sources and speeches from this era, which add another layer of intensity to its activists’ work.

Appendix: Methodology and issues

These visualisations were originally hand-drawn (see below) and then digitally altered and coloured using Sketchbook (iOS).

Top-down photo of a piece of paper containing sketches of the above visualisation, as well as a battered eraser, a micron pen, and a small pencil.
Image by author.

Scaling was difficult with pen and paper, and not made much easier digitally, either. I guess this is something Minard could have complained about, too.

I thought about what colours to use for a while today. I chose colours associated with the SNCC and the modern-day NAACP’s work, including a bright red and green, as well as a muted yellow that ended up looking quite similar to Minard’s original. It’s unfortunate that the colour I used to signify law enforcement was not visually distinguishable from black, because that would have made the message more clear.

With this illustration style, I struggled to draw a line between a final product that looked both professional and inviting (I also struggled to draw lines …). As I got tired, I paid less attention to the fact that my attempt to delineate Alabama’s urban and rural areas began to resemble elementary school sketches. My handwriting is also less legible than I would prefer.

Finally, I would love to find a data source that can tell me exactly how many protesters joined King at every stage of the march so that this viz is less deceiving.

Thank you so much for reading! Any criticism, both of my recounting of this history and of my dataviz methods, is appreciated.



Yaning Wu

she/her. Population Health student @ UCL. Perpetual dataviz nerd. Published on Towards Data Science and UX Collective.