>A Bit of History

> Where did visual representation of information come from? How long has it been around?

\ Otho Mantegazza _ Dataviz for Scientists _ Part 2.1

Wiliam Playfair (1759 – 1823)

In his words, commenting on how data were becoming ever more aboundant in life, and how data visualization (statistical diagrams) could help:

“And it is hoped,that with the assistance of these Charts, such information will be got, without the fatigue and trouble of studying the particulars of which it is composed.”

But…

Maybe we’ve always known how to represent information visually, because that’s the first and most intuitive way to do it.

CURSIVE HIEROGLYPHS

Year: ~1250 BC

Photo: British Museum

Hieroglyphs originated as logographic symbols, and later evolved in syllabic and alphabetic elements, like these ones on the Papyrus of Ani.

BEDOLINA MAP

Year: ~1000 BC

Photo: Luca Giarelli / CC-BY-SA 3.0

The Bedolina Map is one of the most ancient topographical map. It contains around 109 figures of paths, warriors, animals, and wooden huts.

TERRACOTTA LEKYTHOS

Year: ~420 BC

Photo: MET Museum

Mythological stories where told visually on terracotta flasks in Ancient Greece; such as this one, telling of Theseus and Hippolyte fighting Greeks and Amazons

Abstraction

Visual display of information

The graphical display of information, of words, of stories, of topology, has been part of human history as soon as the first pictures on rocks where drawn; and we know that they date back at least 50.000 years, probably more.

Visual Abstraction

But data visualization requires some level of visual abstraction. We are not only drawing what we can see about the world. We are representing data visually, using abstract rules to draw something that did not exist before, and to leverage our visual perception to understand something that wasn’t visual before.

But…

Visual and graphical representation of phenomena and data that are not inherently visual, existed well before the statistical diagrams of William Playfair. Can you think of some?

CLAY MAP OF NIPPUR

Year: ~1400 BC

Photo: Hilprecht, H. V.

In ancient Babylonia, maps were accurately drawn on clay tablets.

Is a map an abstraction? One have to argue that what’s represented in a map is definitely observable in real life, at least in theory. Though, very few of the people drawing maps had the chance to. They relied on measurements and data, representing them on the Cartesian plane.

SONG DINASTY MAP

Year: 1137 AD

Media: Impression of Carved Stone

The Yu Ji Tu (Map of the Tracks of Yu the Great) represents the coastal line and rivers of China in a gridded Cartesian plane, with elevated precision.

MUSICAL NEUMES

Year: ~900 AD

Topic: Gregorian Chant

Musical Neumes are one of the earliest graphical representation of music, drawing over vowels would represent the pitch and modulation in which the text should be sung.

STAFF MUSICAL NOTATION

Year: Current

The staff musical notation, was developed in the last 1000 years. It represents graphically pitch timing and modulation of music, across 5 parallel lines.

Colud it be one of the first way to abstract data and then represent them graphically in a new way?

The Century of Science

In the 17th Century, interest for the natural sciences and disciplines such as astronomy, geography, physics, demographic, grew exponentially.

In that period, data visualization as we know i, was born under the name of “Statistical Diagrams”.

Builing Blocks

Geometry and Coordinates

Analytical Geometry started with the work on the Cartesian System.

In a plane, the cartesian system is a coordinate system that specifies each point univocally as a pair of real numbers, the distances from two perpendicular axes.

It was invented in parallel by Descartes and Fermat in 17th century.

Builing Blocks

Measuring Nature

When people started to look with growing interest to the natural sciences, a question rose: how can we measure things better? How can we make sense of the measurement that we take, which, in a ever changing world, are often, apparently, chaotic?

Probablility and Statistics

Like this, the first statistical approaches were developed by Tycho Brahe to reduce measurement errors in empirical measurements. At the same time, Pascal and Fermat developed the first probability theory.

LONGITUDINAL DISTANCE TOLEDO-ROME

Author: Michael van Langren

Year: 1640

This Longitudinal Distance Chart represents separate measurements of the distance between Toledo and Rome. Each measurement is a labelled point on a x-axis of a one-dimensinal chart. All measurements overestimate the real distance, which is at ~16˚.

This chart contains so many ideas and concepts at once. The idea that measurements contain an error, that they can be repeated and pooled to infer this error, that they can be represented graphically to inspect this error using visual perception.

The x-axis is well defined and labelled, the points show each single measurement, and are labelled also, the title is inserted in the empty space.

SURVIVAL PLOT

Year: 1669

Author: Christiaan Huygens

This density plot shows the age on the x-axis and the survival probability on the y-axis. Huygens estimated Survival probabilities from John Gaunt’s life expectancy data in London.

The graphs doesn’t show data points but an empirical probability density.

WEATHER HISTORY

Year: 1685

Author: Robert Plot

A vertical line plot of the barometric pressure in Oxford for the year 1684.

This 2-dimensional line plot shows the data for each day of one solar year in two lines of facets.

GRAPHIC TABLE

Year: 1737

Author: Johann Heinrich Lambert

The temperature is represented on the x axis, while the y axis represents the month of the year. Lambert binned the data by this two variables, and counted how many days fell into each bin.

Lambert positioned number in space based on the mappings of the x and y axis. Today we are used to map a color intensity to the count numeric value, in what would become a heatmap.

TABLEAU POLÉMETRIQUE

Year: 1782

Author: Charles-René de Fourcroy de Ramecourt.

In this polymetric table the area of each polygon is proportional to the population of the main European cities.

This is one of the first treemaps, in which the area of polygons is mapped to data, and then the polygons are fit in space of the page.

WEATHER MAP

Year: 1816

Author: Heinrich Wilhelm Brandes

The first weather map, showing data collected from 1783. Brandes overlaid isobar lines over the map of Europe.

This map introduces isolines, which are lines that connect points with the same value over a continuous gradient.

CHOROPLETH MAP

Year: 1826

Author: Charles Dupin.

Dupin used shades of dark to overlay the level of education, stratified by region, on the administrative map of France. This approach is now known as choropleth map.

In this choropleth map, the darker the region the higher is the average education level.

The Golden Age

In the 19th century, statistics took a key role in science, economics and policy making.

This is a golden age for data visualization, in which imaginative, beautiful, informative charts were produced to foster progress and decision making.

RADIAL HISTOGRAM

Year: 1858

Author: Florence Nightingale.

Nightingale used this radial stacked histogram to show hospital death causes in the Crimean war of 1854-1855

Aimed a the Queen Victoria and policy makers, this chart showed a peak of deaths from infectious diseases (light blue) during winter months in military hospitals. Florence Nightingale regarded these deaths as preventable, and used this chart to argue for the need of basic sanitation practices, such as air circulation.

COMMEMORATION OF AFRICAN AMERICANS

Year: 1900

Author: W. E. B. Dubois

Time is represented on the y-axis; the size of the path on the x-axis represents populations size. The colors show how to ethnic groups slowly mix together

an American-born Ghanaian sociologist, socialist, historian, and Pan-Africanist civil rights activist. At the universal exposition of Paris, in 1990 he collected photos and charts on the life of African Americans, to challenge stereotypes.

COMMEMORATION OF AFRICAN AMERICANS

Year: 1900

Author: W. E. B. Dubois

Dubois’ work of data visualization for information and storytelling is a great example of clear and easy-to-read charts prepared with an assertive and impactful style.

From 1900 On

In the first half of the 20th century, data visualization was relegated to a secondary role to numerical statistical summaries.

In the second half of the century, instead, John Tukey revived it. Until today, graphics remained one of the main tool in the hand of data analysts.

From 1900 On

Today, with the web based resurgence of graphics and typography, data visualization is a scientific tool, a form of communication and a form of art.

Visual Communication

Visual communication is the oldest form of graphical communication. It predates phonemes and other abstract forms of writing.

Visual communication is a powerful tool for storytelling, it’s intuitive, and it relies on our skills to process broad visual inputs quickly and in parallel.

Since visual communication is intuitive, it delivers (almost) immediate messages. This immediacy makes it potentially misleading. The responsibility for making it clear and not ambiguous is yours (the author).

Data visualization is a part of the broader discipline of visual and graphical communication. Let this whole field inspire you.