In preping for a lecture that I'll give at the end of the month, I've been looking into the more recent history of dataviz - mostly the 1970s. I just realized that celebrating the VISUAL in dataviz is totally a concept that could "only happen now" - a thread:
In the early 20th-century, data visualization rapidly moves away from statistics and towards advertising such as this: "Car sales" - Thomas Cleland, 1924
In the 1930s printing technology continues to be better, cheaper, faster. As a result, designers pull in dataviz techniques for multitudes of publications, which I think is best embodied by Fortune Magazine. "Retaliatory Power" Max Gschwind, 1954
By the 1960's we see dataviz ideas used widely across educational materials of all types - used to make information more understandable and interesting. Many are colorful, illustrated, and all are hand-drawn. "Florida Atlas" Erwin Raisz, 1964
Then something interesting happens in the 1970s. Now rudimentary computers begin to collect data and programmers use those basic machines to then draw the visualizations as well. There are many examples:
In 1975, Edward Tufte taught a statistics course to journalists who were studying economics. These lectures (and others in collaboration with renowned statistician John Tukey) became the foundation for The Visual Display of Quantitative Information. Its cover is... minimal.
Despite my deep misgivings on the interpretations of this landmark book, it has a singular focus on the use of dataviz to visually explore data. This is literally reinforced on each page and is typified by this chart by Paul and John Tukey, 1981.
What we see is a deep disregard for the graphics used in journalism. While this is warranted criticism, his tone throughout the book is that dataviz is best when focused on the fastest reading of the data, and any distraction from this is as best "junk" and at worst unethical.
So, here's where things get interesting. Human Computer Interaction (HCI) becomes a thing. Compters are fun, they are intuitive, and we develop a relationship to them. Professional artists still draw amazing dataviz designs, but more and more people start clicking on computers.
The previous tweet was from a 1985 Apple IIe, and this is from 1992, the Amiga Workbench 3. We see massive advances in color and interface design. Throughout the interface we see icon design and informational systems. Our tools *become* a dataviz of how we are using them.
About the same time (80's, 90's) we also see the rise of the "Graphics desk" at a few newspapers/mags. This work ranges from Nigel Holmes illustrated charts and Peter Sullivan's investigative diagrams to John Grimwade's scientific illustrations (below is from Sullivan, mid-80s)
Computers continue to be refined. Powerpoint is invented in 1989. Photoshop and HTML are both invented in 1990. CSS in 1994, Javascript in 1995. Foundational concepts are established to meet a tacit need for creating visual designs digitally.
The iPhone is released in 2008. It is literally a revolution felt as much culturally as technologically. What was HCI becomes User Experience (UX) Design. Computer culture is reinvented/merged with business culture. Digital tools are full of icons, every interface is a system.
Digital photography and data visualization are central to how the technology is presented and marketed. Our lives become more further linked to our technology and our data. Our lives are accompanied by the digital dataviz twin in our pockets.
In 2011, D3 and Bootstrap are both introduced. D3's computational drawing makes it the foundation for many viz platforms while Bootstrap's broad adoption becomes a defacto template for web design. Tools are democratized widely and further standards are introduced.
Then humans do what humans do - they color outside the lines. Dataviz becomes even more human and stories become the focus of our communications. Data journalism leverages aspects of scientific illustration, Mike Bostock, co-inventor of D3, is at the NYT & boundaries are pushed.
So here we are ~10 years later. We focus on data ethics and literacy, employ amazing tools, and leverage best practices. Data humanism has become a truism. Data art has effectively merged with cryptocurrency (well, for the moment). Our tools are nuanced, malleable, and personal.
>THIS< is why we are publishing Nightingale in print! THIS history, THIS moment in time has to be documented (yes, we're not the only ones ). The creation of a community-oriented magazine like this could not have happened at a better time in its history: nightingaledvs.com/subscribe/
If you agree with me, then please do subscribe to Nightingale Magazine. We call it a "collectible community celebration" and this is exactly what we aim to make it - a celebration of all the various facets of what dataviz can be today and tomorrow. /end
Tagging you @mbostock out of respect for your place in this timeline! 🍻
@nigelblue tagging you on this thread as well. Please jump in and tell me what I got wrong! ha! Thanks! 🍻
Tagging @EdwardTufte out of respect to your amazing scholarship (albeit pretty grumpy, haha) 🍻
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Monroe N. Work was an African American sociologist, scholar, and researcher who spent his life collecting information and helping others to understand it. The highlight of his career, according to Work, was the nine editions of the Negro Year Book between 1912 and 1938.
He collaborated with W.E.B. Du Bois and Booker T. Washington, positioning him at the intersection of Black leadership and education in the US for most of his life. He started the Dept. of Records and Research at the Tuskegee (@TuskegeeUniv) to collect facts:
I lecture on topics related to dataviz, but last week I presented this at @InfographicsNL on my evolving thoughts on @midjourney (+ @OpenAI & @StabilityAI). I have a few interesting ideas on how we can use language and images to design AI pictures /🧵
The talk does 2 or 3 things. First, it explains how to use Midjourney. This is a diagram of the variations and upscaled images and shows how quickly ideas can morph:
The lecture also walks through my story of using these new tools. When I started I was focused exclusively on using Midjourney to explore new design patterns in Isotype research.
People are loving AI-fueled dataviz, so here's a thread on a project that I made in May but never shared. The prompt was "hyperdetailed illustration of economic prosperity, 1965, lenticular rainbow, by isotype". The design put this among the first responses:
A few iterations later, all of a sudden a bunch of people show up at the bottom. The rainbows begin to curve and the city reciedes.
Among the following versions, was this one pulling the bars into buildings. The texture at the bottom may have been the people?
Today we launched a novel dataviz project on a serious subject - the mental health of working parents. Our team (led by @adsevenfour) felt that the human side of the data really needed to be stressed - what better way than with 100 little people walking between the categories! /1
You can really explore the data in a variety of ways and view some demographic cuts as well
One of the most interesting findings is in the days off work, as most people didn't take more than 5 days off.
I went to that Van Gogh immersive thing today and it was super interesting. Many thoughts:
First off, it is not narrated, and the paintings are treated more like a visual remix. Elements if multiple paintings are collaged in a single animated scene
As you prob know, audiences are flocking to the show(s) and I wonder if it is because the visuals use the language of film rather than the language of painting.
I haven’t been sharing as much lately because I’ve been trying not to buy too many more books, but today I started looking through Karl G Karsten’s fantastic 1926 Charts and Graphs. Here’s what we says about our favorite chart type:
I really like this pile of shells:
I have no idea what this is, but I feel like a drew a sketch of this earlier this afternoon!