Historical Viz Digest: Issue 2
March 2019: Trees, Rubber, and Yukaghir diagrams
The Data Visualization Society (DVS) historic-viz channel is a non-stop thread of amazing historical visualizations from before 1983 — the landmark release date of Tufte’s seminal Visual Display of Quantitative Information. There are multiple new posts a day and the commentary is nothing short of thrilling. Here’s a collection of posts with some commentary from the people who posted about them.
One of the most remarkable aspects of this vibrant community is the disparate fields from which these visualizations are posted. Science, music theory, anthropology, climatology, and business (just to name a few) have been presented this past month. Whole fields of thought are shared hour by hour and we’re all happy to have the opportunity to learn more about each in due passing. Here are a few highlights from the past few weeks:
Max Furbringer: ‘Phylogenetic Tree of Birds’, 1888.
Karol Stopyra wrote: I like this one as it links two hierarchical layouts: tree and pack layout (3 side diagrams). As the description says: “plates […] show horizontal projections for the upper middle and lower section [of the tree]”. The ‘bubbles’ (bird species) are grouped (outlined) together to create something we could name today as a ‘force directed graph’ where the force is common ancestry. Screenshot comes from ‘Design for Information’ by Isabel Meirelles.
Then, Will Chase leaned way in and provided a number of further illustrations showing additional phylogenetic trees. He was kind enough to summarize them here:
“In 1837, Charles Darwin revolutionized the way we understand and visualize evolution with a simple sketch in his journal that he titled: “I think”. Several subsequent examples came in the late 1800s from naturalist and illustrator Ernst Haeckel. What followed was a rapid adoption and refinement of Darwin’s design, which would come to be known as a phylogenetic tree. A phylogenetic tree is a bifurcating graph where each terminal node (a “leaf”) represents an extant species, while each interior node (just called “nodes”) represents an extinct ancestral species. The leaves are clustered by species similarity, which is determined in modern trees by complex algorithms that incorporate evolutionary theory and statistical computing to cluster species based on their DNA (historically, species were clustered based on morphological similarities). The edges of the graph (called “branches”) are often length-scaled to some parameter like mutation rate (the “speed” of evolution) or time.
“Modern phylogenetic trees often suffer from data overload, trying to visualize hundreds or thousands of species simultaneously. Visualizing large trees is tricky because you quickly end up with visual overload and a hairball style graph, so it’s important to consider layout, coloring, highlighting, and grouping. While we have a long ways to go in defining best practices for phylogeny visualization, some standout examples have provided stunningly beautiful and insightful visualizations. For instance, in 2016, a group of researchers led by Laura Hug and Jillian Banfield published a “New View of the Tree of Life”, which in addition to being a gorgeous figure, perfectly illustrated several recent developments in our understanding of biodiversity. If you’re looking for humans, you won’t find them here, we’re so insignificant that we don’t even get a label. In fact, the entire animal kingdom is found in green at the bottom, tucked within the tiny “Opisthokonta” branch. Much like NASA’s famous “pale blue dot” photograph of earth, this tree gives an imposing and humbling sense of scale to humanity, showing how utterly dwarfed we are by the myriad bacteria and other organisms that populate our planet.”
Ladislav Sutnar “World rubber has the bends”, 1946
Our new co-moderator Stephanie Tuerk posted up this gem. “One of the places I wound up was looking at the work of the Czech graphic designer Ladislav Sutnar. (I got to him by remembering the work of architect/proto-information designer Knud Lonberg Holm to reconceive the architectural product catalog Sweets in the 1940s — which was done in conjunction with Sutnar.) I would say that most of his work is organizing information through graphics rather than data visualization per se, but here is a layout he did for Fortune magazine in March of 1946 that is both.
I think that the isometrically projected chart is hard to read, but what I appreciate about this and something that many 1940s-era data viz type things have in common, is the interconnectedness of all of the elements on the page. I would argue reflects ideas of the total work of art and the lack of division between artistic disciplines that came out of both the Bauhaus and Soviet Constructivism. You can read more about Sutnar here: https://www.aiga.org/medalist-ladislavsutnar (edited)
Sutnar was quoted where he describes a conscious shift away from 2D charts in which “color was used only to facilitate a better understanding,” to what he called a “dramatized” approach that emphasized three-dimensionality and (my interpretation) a greater unity among elements, in which the emphasis shifted from the representation of numbers to the representation of concepts/ideas. (i.e. from data viz to illustration) Maybe this was his attempt to do the latter (3D, everything unified) with the former?
A. Keith Johnston, “Illustrations of Chartography”, 1852
A combination of illustrative zeal and metric definition, this is the frontispiece to the “Atlas Of Physical Geography” by A. Keith Johnston in 1852. In the introduction to the book Johnston “lays claim to a level of accuracy and degree of information not hitherto attempted in educational works.” This image is part scale, legend, and artistic representation to be elaborated on in many of the other charts in this graphic-rich catalog.
Cameron Yick commented: “I like the multi-slice cutaway views of the mountains in the middle. I haven’t seen cutaways outside of mechanical engineering diagrams (for machined parts) or architectural diagrams.”
Zooming in closer, I find it interesting that Johnston includes the “columnar” and “terraced” measurements in the same illustration as mountains and volcanos. The mixture of scientific illustration and charting diagram is both unique and very engaging. I especially love the central illustration below, which shows a host of natural phenomena in the same image with descriptions following along at the bottom denoted by flying bird shapes!
Émile Cheysson, “Recettes brutes des théâtres et spectacles de Paris de 1878"
The amazing RJ Andrews then posted this winner. “THÉÂTRES DE PARIS was designed by Émile Cheysson for the “Album de Statistique Graphique de 1889”. Pictured is a detail of a wide composition. Each theater is represented by a “fan diagram” with one wedge for each year. The total area of each wedge corresponds to that year’s gross receipts. Years of the Universal Exposition, which often attracted additional business, are highlighted in yellow.”
RJ continues “I love so much about this: Its semitransparent yet bold color across a beautiful base map. How the fan diagram mimics the seating layout of the theaters. How macro comparisons between theaters (the scale of receipts is consistent throughout) is just as inviting as microscale trends across a single theater’s years (the dashed line is a summary average of just the years between the Expositions).”
Social network visualizations from the Yukaghir people, a Siberian tribe. Discovered in 1895.
Pierre Dragicevic added this social network visualizations from the Yukaghir people, a Siberian tribe, whose first written account dates back to 1895. It’s from Margaret Mead’s, “Golden Age of American Anthropology” from 1960.
“I think this has all the elements of a data visualization: 1) it represents data cases (which are of two types here: people and relationships); 2) it maps data cases to visual marks (nodes for people and links for relationships, just like a standard network visualization but with 1-D layout; 3) it maps data attributes to visual variables (both on nodes and on links). I don’t know to what extent the visual mappings are codified, but from the book, they seem pretty codified.
“I agree most social network visualizations don’t depict the kind of relationships depicted here! But I think for most social network analysts, a social network is any network where nodes are people.
“If you’re interested in learning more, I suggest “Visible Speech: The Diverse Oneness of Writing Systems” by John DeFrancis. It provides important contextual information that the original book doesn’t have.
I found this accidentally a few years ago by doing a literature search about writing systems, and I’m pretty sure few visualization people have seen this before and I didn’t know what to do of it, so I thought I would post it here.”
We’re certainly glad you did, Pierre!
Francisco de Salinas, “De Musica Libri Septem”, 1577
Petra Isenberg shared this gem. While she didn't know much about the image, it bears further investigation.
The diagram is from De Salina’s 1577 book “De Musica Libri Septem” where he outlines the musical temperament, known as “meantone” (or midtone) which Wikipedia explains to us that it is “obtained by slightly compromising the fifths in order to improve the thirds. Meantone temperaments are constructed the same way as Pythagorean tuning, as a stack of equal fifths, but in meantone, each fifth is narrow compared to the perfect fifth of ratio 3:2.”
Francisco de Salinas (1513–1590) has one heck of a story too. He was a Spanish music theorist and organist who went blind at the age of 10 but went on to study the humanities, singing, and organ at Salamanca University. He then moved to Rome and became the organist in the chapel of the governor of Naples, the Duke of Alba. He returned to Spain around 1559 and became a professor at Salamanca University where he wrote his treatise “De Musica Libri Septem” in Latin.
Salinas and Gioseffo Zarlino are considered to be the first to describe “midtoning” in mathematically precise terms. He describes three types of midtones that he considers suitable for keyboard instruments and remarked that it is “languid” but not “offensive to the ear”.
Willard Brinton’s “Checking list for Graphic Presentations” from his book Graphic Method (1920 edition)
Let’s close out this article with a checklist from Willard Brinton’s Graphic Method (1920 edition. Francois Dion shared these 2 pages and elaborates “The 1939 edition is actually quite a rewrite. If Graphic Method was hard to find, Graphic Presentation is harder. None of the reproductions are in color, which is a real shame.”
What I find amazing about this list is Brinton’s attention to the details of hand-crafting visualizations at this time. He elaborates on methods of hand-drawing and photographically reproducing charts and even includes detailed specs for using Ben Day dots (what we call a halftone screen) as well. Of course, the basic rules for visualization are also included (“Are all dates accurately shown”) and in the following paragraph Brinton elaborates: “Though graphic presentations are used to a very large extent to-day there are at present no standard rules by which the person preparing a chart may know that he is following good practice.”
This list of rules is located at the very end of the book. A page later, Brinton concludes the book with a sort of value proposition for data visualization standards. It’s an appeal that I find to be both scientific and warmly humanistic.
Ok, that it’s it for this month! Thanks to everyone in the #Historic-Viz channel! We’ll be posting more regularly so please follow The DVS publication to get more stories right in your inbox!
Here are some more articles in the series:
The Data Visualization Society is developing meaningful resources in order to establish a discourse and remove barriers between practitioners across tools and industries. To sign up please register at datavisualizationsociety.com/join
