The History of Data Visualizations - From Cave Drawings to Tableau

This paper traces the history of visual representation, from early cave drawings through the computer revolution and the launch of Tableau. We will discuss some of the pioneers in data research and show how their work helped to revolutionize visualization techniques. We will also examine the different styles of data visuals, discuss some of the barriers to making effective visuals and the methods we use to overcome those barriers. In the end, we will show the power (and limits) of human perception, and how we can use data to tell stories – much like those of the earliest cave drawings.

Visualizing the Past

Visual representations are a fundamental component of human learning and understanding. To study the impact and evolution of data visualization, we must first look to the past and see how our ways of shaping and representing data have changed over time. As we will see, mankind has used visualizations to instruct, convey meaning and tell stories since the dawn of time. Maps, in particular, have a long history.

Different Data Visuals for Different Needs

There are two common types of visual representations of data. Both are very important and both have different requirements when it comes to designing great visualizations.

1. Presentation - Uses data visuals to communicate. This type of visual representation has two roles: a presenter and an audience.
2. Visualization - This is a fairly new term and the idea is to use visuals to think. Here, the experience is active and involves people trying to answer questions.

1700-1900: Visualization is Transformed

A key point in the history of visualization is with William Playfair, a Scottish engineer who is widely regarded as the father of statistical presentation. Playfair published a book in 1786 called the Commercial and Political Atlas which used graphical representations of data to describe England’s balance of trade. Many of Playfair’s innovative data visualizations are still in use today.

It wasn’t long before statistical graphics were being used for presentation. One famous example comes from Dr. John Snow, a British physician who used statistical graphics to deal with London’s cholera epidemic of 1855. Snow plotted individual cases of cholera as dots on a map of London. These dots showed that the majority of cases could be traced to a water pump on Broad Street. An investigation of outlying cases showed they, too, had connections to the Broad Street pump. Snow removed the handle from the contaminated pump and the cholera epidemic subsided. This shows how the power of visualization can answer questions and, in this case, even work for the public good. Snow’s map also works as an effective example of the Presentation style; Snow’s data was strong enough to persuade city officials to remove the infected handle and quell the outbreak.

The 20th Century Brings Advancements and Abuse

By the mid-1900s, statistical graphics had grown in both popularity and abuse. This abuse prompted American writer Darrell Huff to publish How to Lie with Statistics in 1955.

More than a decade later, French theorist Jacques Bertin published the Semiology of Graphics. Bertin was particularly interested in statistical mapping and he observed that data views involved three types of marks: Points, Lines and Areas. Those marks had certain properties, the most important of which was Position. Another six properties independent of Position include Color, Size, Shape, Gray, Orientation and Texture. Bertin turned these marks and properties into guidance on how to design graphical data.

To learn more about the history of data visualization, read our whitepaper.

This paper reviews some simple lessons that can help us to tell effective stories including:

• Trust is a key design issue
• Visuals must be expressive and convey data accurately
• Make Effective use of your visuals by exploiting human perception
• The importance of context to help your audience to better understand your data view

Always strive to tell stories with your data and your visuals understand that good stories involve more than just data.