The History of Visual Communication

Long before there were "data visualizations," there were pictures that carried information. The story of data visualization is, in the end, the story of one of humanity's oldest habits: turning abstract things into shapes we can see.

This page is a short, deliberate detour through that history. Knowing where charts come from will change how you read every chart for the rest of your life.

Before there were charts, there were maps

The oldest known map is a star chart carved into a mammoth tusk roughly 32,000 years old, found in a cave in Germany. The earliest geographic maps we have come from Babylon around 600 BCE — clay tablets that show rivers, cities, and the boundaries of the world as the mapmaker understood it.

Maps are interesting because they were the first widely-used visual abstractions. A line on a map is not a river — it is a symbol for a river. The map and the world are not the same thing. Understanding this gap — between the thing and the picture of the thing — is the entire foundation of data visualization.

The first statistical graphic

In 1644, the Flemish astronomer Michael van Langren made what historians now consider the first statistical chart. He was trying to figure out the longitude between Toledo and Rome — a fierce open problem at the time. Twelve different astronomers had published estimates ranging from 17° to 30°. Van Langren drew them on a horizontal line:

Toledo                                          Rome
  |                                              |
  | x   x x  x  x x   x x x      x  x       x
  17°  18°  19°  20°  21°  22°  23°  24°  ...  30°

Just by seeing the dots, you immediately notice they cluster on the left side. The eye does the statistics for you. Van Langren had discovered something profound: a picture of disagreement is often more honest than a single number.

Playfair and the birth of the chart

If van Langren is the great-grandfather of statistical graphics, the father is William Playfair — a Scottish engineer and political economist working in the 1780s. In a single book, his Commercial and Political Atlas (1786), Playfair invented or popularized three chart types we still use today:

• The line chart, for showing change over time.
• The bar chart, for comparing categories.
• A few years later, the pie chart, for showing parts of a whole.

Playfair's leap was philosophical. He argued that economic and political data could be seen, not just read. To a generation raised on tables of numbers, this was strange and a little suspicious. A bar chart was a picture of trade — was that serious? Today the answer is obvious; in 1786 it was not.

Why pictures became scientific

Three things had to come together for the chart to become a standard scientific tool:

1. Cheap reproduction. Printing presses and engraving had to become good enough to produce a clean chart at low cost.
2. A culture of evidence. Scientific journals had to start expecting visual evidence rather than narrative description.
3. A theory of perception. Researchers had to start asking why a chart works at all — why a length is easier to compare than a shaded area, why color carries meaning, and so on.

We will spend an entire chapter on the third one later in this course. For now, sit with this thought: every chart you see today is the descendant of a 250-year-old idea that pictures of data are a legitimate way to know things.

Why this matters for you

You are about to write Python code that produces charts. The code is the easy part. The hard part is the same hard part Playfair faced: deciding what to draw, and why anyone should believe it.

When you make a bar chart in Plotly Express, you are standing on two and a half centuries of cumulative human invention. Treat each chart you produce as a small act of communication, not a piece of decoration.

Check your understanding