I came across this gem – a book on data graphics that is almost 100 years old called Graphic Methods for Presenting Facts by Willard Cope Brinton, written in 1914. It’s the “Ghost of Dataviz Past”, from a time when data graphics were made by hand with paper, ink, and a draftsman’s tools rather than a slick software program designed to “wow” at the click of a mouse button.
I began reading through this ancient tome wondering whether I would find it to be an obsolete and amusing relic of a less sophisticated era, or whether I would discover timeless wisdom and long lost, often neglected advice. I was pretty sure it would be the latter, and I was right. The issues Brinton and his contemporaries faced are the same we face today.
Here are a few examples of the advice Brinton gives to those who would design data graphics. His words are eerily relevant.
On the Real-World Importance of Visualizing Data
Brinton stressed the critical importance of getting data visualization right. He wrote his book primarily for businessmen, who, he believed, should be well-versed in the graphical methods of comparison (the book was put together, in part, from a series of lectures delivered to Harvard Business School students).
“As the cathedral is to its foundation so is an effective presentation of facts to the data” Brinton, p.2
“One of a business man’s chief assets is his ability to show things to others in their true proportions.” Brinton, p.20
Brinton observed that one of the chief problems in the field of data visualization was the lack of universally accepted standards, and felt that such a system would soon emerge and would be an international language like mathematics:
“The trouble at present is that there are no standards by which graphic presentations can be prepared in accordance with definite rules so that their interpretation by the reader may be both rapid and accurate. It is certain that there will evolve for methods of graphic presentation a few useful and definite rules which will correspond with the rules of grammar for the spoken and written language” Brinton, p.3
One could argue that many useful rules have indeed been put forward by Brinton and others since his time, but the question is whether or not these rules have been widely adopted.
Let’s see for ourselves:
On Pie Charts
Brinton used the term “sector method” to describe pie charts, and I was not surprised to find that he was opposed to their usage (John Tukey would later famously proclaim, “There is no data that can be displayed in a pie chart, that cannot be displayed BETTER in some other type of chart.”)
Regarding the above “sector method” chart, Brinton had the following to say:
“Fig 2 is a form of chart used probably more widely than any other form to show component parts. The circle with sectors is not a desirable form of presentation however because it does not have nearly such flexibility as the method shown in Fig 1″ Brinton, p.6 (Fig 1 is a horizontal bar broken into sections)
He goes on to discuss various drawbacks to the “sectional method”: difficulty arranging the names of the different components, the “impossibility of placing figures in such a manner that they can be easily compared or added” and more.
What is remarkable is that in spite of this and continued criticism of the pie chart, it remains “perhaps the most widely used statistical chart in the business world and the mass media” (quote from Wikipedia)
On Embellishment and Aesthetics – Infographics
The past few years have seen an explosion of over-embellished (read: confusing) bad “infographics”, and an associated (and understandable) backlash against all attempts to make any chart look “beautiful” in order to catch a reader’s attention.
It seems these attempts at entertainment with data were going on at the turn of the previous century, and Brinton gave what I feel is very balanced advice regarding the use of what he called “eye catchers”. If only every eager infographic designer internalized these examples and gave a “Hippocratic Oath” not to do harm with data in such ways:
1. How NOT to do it:
“This particular diagram contains nothing by which accurate comparison may be made. No figures are given and it is impossible to tell whether the different money bags should be compared on the basis of diameter, area, or volume.” Brinton, p.21
“Fig 18 gives a statement which the illustration does not support. In the first place the dates of the two years compared are not given. In the second place it is impossible for the reader to tell whether the diagram is drawn on the basis of one dimension two dimensions or three dimensions. It would be a hopeless task to fit the area of the smaller washing machine into the area of the larger washing machine. Methods like this cannot be too severely condemned.“ Brinton, p.21 (emphasis mine)
And, at long last, Brinton unleashes this diatribe:
So are all embellishments bad then?
In spite of this clear bias toward clarity in data graphics, Brinton does not universally condemn attempts at embellishment. In particular, he finds the following potential benefits:
“Nevertheless the cartoonist style should not be broadly condemned, for it has tremendous possibilities. It is possible to combine the cartoonist’s wonderful power of arousing interest with methods of presenting facts which will give a numerical interpretation that cannot be misunderstood. There is a great opportunity waiting for the man who can combine cartoon methods with accuracy of numerical statement.” Brinton, p.21
Some more tolerable examples:
1. Use of a steamboat image at the top of the chart to attract attention:
2. Use of “icons” to the left of the rows to add a graphical label:
3. Finally, a comparison of how not to do it (first) along with a more tolerable example (second):
NOT Okay:
Okay:
What’s the Overall Message?
So Brinton was in favor of tasteful uses of images and embellishments, just as long as images aren’t used to misrepresent the actual data. Accurate depictions of proportion must be preserved.
I find it fascinating that the world 100 years ago had this book, and yet today we are repeating the sins that were clearly spelled out in it. We have seen such incredible advances in the tools used to create data graphics since Brinton’s time, but not in the corresponding methods and practices.
An Evolution Coming?
Brinton predicted the emergence of a set of standards, a universal language of data visualization, but such a system has not been widely accepted as far as I can tell. There are pockets of practitioners that adhere to certain mutually enforced practices, but the vast majority rely on poor software defaults and continue clicking on every titillating infographic that pops up on their screen.
Is there a way to make Brinton’s vision a reality, or is it just a Utopian dream?
Next: Brinton on Principles of Presenting Data
Pingback: Paletleme Amirliği, 8 Temmuz 2012 « Emrah Göker'in İstifhanesi
Wow, this blows my mind. It’s incredibly interesting to see old school BI/Viz from so long ago. Had no idea anyone had thought so far ahead back then.
Awesome post Ben!
Hey Will! I know, amazing to think they were wrestling with the same kinds of issues a century ago! Thanks for the comment, and it was great meeting you today.