ChapterPDF Available

Introduction to Playfair's Commercial and Political Atlas and Statistical Breviary

Authors:
  • Independent Statistician and Author

Figures

Content may be subject to copyright.
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Introduction
Sometime in 1787, just two years beforeFrance was plunged into revolu-
tion and chaos, the Count of Vergennes delivered a package to the royal
court of France for the attention of the king. The gift for Louis XVI had
come to Vergennes from Lord Lansdowne, an English politician who was
on intimate terms with many in the upper echelons of Parisian society.
Vergennes was certain that Louis XVI would be very interested in the
contents of the package.
The gift was a book written by a young Scottish engineer and en-
trepreneur who had recently moved to Paris with hopes of making his
fortune. His book had been published in London during the previous
year and was entitled The Commercial and Political Atlas but, unlike more
conventional atlases in this era of great exploration, it contained no maps.
It did contain charts, but of a new and unfamiliar variety. Louis XVI,
an amateur of geography and the owner of many fine atlases, examined
his acquisition with great interest. Although the charts were novel, Louis
had no difficulty in grasping their purpose. Many years later, their author
wrote that
[the king] at once understood the charts and was highly pleased.
He said they spoke all languages and were very clear and easily
understood. (Playfair, 1822–3)
Afurther indication of the king’s approval was the royal permit he
granted for the establishment of a factory to work metals in Paris.
Playfair had intended to use a steam engine to drive a rolling mill,
modeled on the machinery and practices in the Birmingham factory
1
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
2Introduction
of Boulton & Watt, where he had worked from 1777 until 1781. In
addition to his endorsement of the venture, Louis XVI donated the
land.
The Atlas that had captured the king’s imagination contained nu-
merous tables and graphs that summarized trade between England and
several other countries, as well as a variety of charts that displayed eco-
nomic data. In total, the volume contained 44 charts. The king of France
had grasped fully the significance and utility of the novel representa-
tions in the Atlas.Importantly, he had understood the universal appeal
of the new diagrams. Of course, the use of tables to present economic
data was not new, having been common for more than a century after
John Graunt (1620–74), who had used them extensively in his Natu-
ral and Political Observations Made upon the Bills of Mortality, and
SirWilliam Petty (1623–87), who had examined the role of the state
in the economy in his Treatise on Taxes and Contributions.Coinciden-
tally, both books were published in the same year, 1662. But the pictorial
representation of statistical data was revolutionary. The Atlas showed,
for the first time, how economic data could be represented by charts.
The favorable assessment of the ill-fated Louis XVI – who was to per-
ish under the guillotine less than six years later – was both fitting and
prescient. A century and a half later, in 1937, the great American his-
torian of statistics, H. G. Funkhouser, echoed the sentiment of the king
when he said that “the graphic method is rapidly becoming a universal
language.”
To day there is scarcely a field of human activity that does not make
use of statistical charts like those in the volume delivered to the king of
France. The invention can lay fair claim to being one of the most versatile
and useful tools for analyzing and displaying data in the sciences and
humanities, in commerce and the arts, and in everyday activities that
affect us all. Graphs convey comparative information in ways that no
tables of numbers or written accounts ever could. Trends, differences,
and associations are seen in the blink of an eye. The eye perceives instantly
what the brain would take seconds or minutes to infer from a table
of numbers, and this is what makes graphs so attractive to scientists,
business persons, and many others. The charts allow the numbers to
speak to all, and they transcend national boundaries – a Chinese can
read the same graph that a Russian draws. There is no other form of
human communication that more appropriately deserves the description
universal language.”
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Introduction 3
WILLIAM PLAYFAIR
The author of the Atlas was no ivory-towered theoretician. William
Playfair was trained as a practical engineer by giants of the Industrial
Revolution. Although a craftsman by trade, he was also exposed to the
best academic minds of the Scottish Enlightenment, which has so pro-
foundly helped to shape our modernworld (Broadie, 2003; Buchan,
2003; Herman, 2001). William was born on 22 September 1759, in the
small rural village of Liff, near the city of Dundee. He was born a twin,
the fourth child, in the family of the Reverend James Playfair, a Presbyte-
rian minister of the Church of Scotland. Sadly, the twin brother, Charles,
like too many other children in those days, did not survive to see his
first birthday. In the early years, the Playfair children were educated at
home by their father. However, upon the relatively early death of James
Playfair, when William was just 12, the role of teacher was thrust upon the
eldest brother John, then 24. John would soon gain worldwide fame as
amathematician, physicist, and geologist and would become one of the
most distinguished professors at the University of Edinburgh. William
Playfair was raised and educated in the presence of genius.
John’s scientific approach was unequivocally empirical; one task that
he gave his younger brother was to keep a graphical record of daily
temperatures. Many years later William acknowledged this childhood
exercise as the inspiration for his economic time series line chart. Addi-
tionally, and also significant for his intellectual development, John intro-
duced William to many of the great figures of the Scottish Enlightenment,
such as the philosopher Dugald Stewart and the economist Adam Smith.
John also commended his brother to William and Robert Small, educa-
tors who were exceedingly well connected in the 18th-century world of
letters, science, medicine, and politics. The Small brothers would play a
crucial role in the future training of William Playfair.
The Rev. Robert Small and the Rev. James Playfair were well ac-
quainted, being fellow ministers in nearby parishes in the Presbytery of
the city of Dundee. Both had received their Doctor of Divinity degrees
from St. Andrews University, the oldest in Scotland, and both were enthu-
siastic teachers; they had many interests in common. Robert’s brother,
Dr. William Small, was trained as a natural philosopher and physician
at Marischal College, Aberdeen. In 1758 he joined the faculty of William
and Mary College in Williamsburg, Virginia, where he served as a pro-
fessor of mathematics and natural philosophy for six years. By Thomas
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
4Introduction
Jefferson’s own admission, William Small was his most important men-
toratWilliam and Mary, and the pair maintained a close private friend-
ship and correspondence through the difficult years surrounding the
American Revolution. In his autobiography (1821), Jefferson wrote:
It was my great good fortune, and what probably fixed the des-
tinies of my life that Dr. Wm. Small of Scotland was then professor
of Mathematics, a man profound in most of the useful branches of
science, with a happy talent of communication, correct and gen-
tlemanly manners, and an enlarged and liberal mind. He, most
happily for me, became soon attached to me and made me his
daily companion when not engaged in the school; and from his
conversation I got my first views of the expansion of science and of
the system of things in which we are placed. (4)
On his return to Britain, William Small became a founding member
of the Lunar Society of Birmingham (Schofield, 1963; Uglow, 2002). In
many ways, he was the central figure of the group, and his early death at
41 caused great distress to James Watt and the other members.
At the age of 14, William Playfair left the family home to apprentice
with Andrew Meikle, a well-known Scottish engineer and the inven-
torofanearly threshing machine. Meikle was miller and millwright to
the Rennie family, owners of the Phantassie estate at East Linton, near
Edinburgh. One of the Rennie boys, John, also worked at the Houston
Mill under Andrew Meikle’s instruction during the same years that
William Playfair served his apprenticeship. John Rennie would later
become the renowned engineer responsible for the London, Waterloo,
and Southwark Bridges, as well as several other significant engineering
structures. After three years with Meikle, William was recommended by
Robert Small to the position of draftsman and assistant to James Watt,
during the early days of the Birmingham steam engine factory.
James Watt (1736–1819) ranks among the most famous of all engi-
neers. This consummate craftsman and scientist did not build the first
steam engine, as is so often popularly supposed, but there is no doubt
that his improvements converted a primitive, balky, awkward, and inef-
ficient device into the workhorse of the Industrial Revolution. His devel-
opment of the Newcomen engine was so successful that for all practical
purposes we may say that Watt did “invent” the steam engine. His most
important contribution, in 1765, was the separate condenser, which he
included in his first patent of 1769. The work was largely completed at
the University of Glasgow, but it was not until 1776 that the first practical
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Introduction 5
engine was built, and the construction of such engines did not become
routine until the mid-1780s. Had it been left to Watt, a man subject
to despondencies, his momentous ideas might never have come to full
fruition and achieved great commercial success. The eventual success was
based on Watt’s collaboration, starting in 1774, with Matthew Boulton,
who had established, in the Midlands city of Birmingham, an engineer-
ing factory called the Soho Manufactory that became world famous for
its organization and novel equipment. It was William Small who intro-
duced Boulton to Watt and who had encouraged the partnership that
led to the development of the steam engine manufacturing company,
Boulton & Watt, which was to revolutionize work throughout the world
in the most fundamental way.
William Playfair arrived in Birmingham, England, in 1777. He worked
in a variety of capacities at Boulton & Watt, but one of his most impor-
tant duties was as draftsman and clerk to James Watt himself. Watt did
not spend much time at the Soho Manufactory, preferring to work alone
in his house at Harper’s Hill. It was there that Playfair helped Watt with
his engineering drawings, although Watt, who was always a demand-
ing critic, did not have the highest opinion of Playfair’s drafting skills,
referring to him as a “blunderer” in a letter to Boulton in 1778. When
it came to patent applications Watt prepared his own drawings, appar-
ently because he was less than pleased with Playfair’s efforts. Nonetheless,
Playfair continued in this post until the autumn of 1781, so Watt cannot
have been completely dissatisfied with his work. Indeed, Boulton indi-
cated, in a letter to Watt, that he was sorry that Playfair was leaving, since
Watt would no longer have his assistance in drafting. Blunderer or not,
Playfair’s experience in drafting and printing drawings for Watt would
later serve him well when he turned his hand to writing.
During his time in Birmingham, Playfair became acquainted with
several members of the Lunar Society. This distinguished group of busi-
nessmen and scientists included Boulton, Erasmus Darwin, Edgeworth,
Keir, Priestley, Watt, and Wedgwood. The unusual name of the society
derived from the meeting time of the group – they met monthly, from
1765 until 1813, on the Monday evening closest to the full moon so that
there would be sufficient light for the late night walk home. The Lunar
Society was second only to the Royal Society as an important meeting
place for scientists and inventors. Its members were interested in more
than pure science – they were passionately engaged in the application of
new ideas in natural philosophy to manufacturing, mining, transporta-
tion, medicine, and education. The members of the Lunar Society were
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
6Introduction
the heart and soul of the Industrial Revolution, and they were convinced –
with much justification – that they were changing the world for the better.
Thus William Playfair was privileged to be at the cutting edge of science
and industry in a Britain that was to dominate the world in the century
following the Industrial Revolution–arevolutionthat was spawned in
late 18th-century Birmingham. Playfair rubbed shoulders with the lead-
ing figures of the day in science, engineering, business, and politics, and
they, unknowingly, helped to shape his statistical creations.
In 1779, while still at Boulton & Watt, Playfair married Mary Morris,
and in 1780 their first child, John, was born. A draftsman’s wages may not
have seemed adequate to support the new family, and as soon as Playfair
felt he had learned enough to strike out on his own, he left Boulton
&Watt in 1781 with a fellow Soho employee, William Wilson, to form
asilversmithing business in Marylebone, London. From the start, the
new venture was plagued by disputation and bad debts. In a pattern
to be repeated many times in the coming years, Playfair had embarked
upon a speculative grand scheme that was doomed to failure. It seems
that his reach always exceeded his grasp. Despite obtaining four patents
for devices to fashion metal objects, from silver trays to horseshoes, the
business was not successful and Playfair turned his hand to writing.
Playfair’s developing interest in writing about economics was in-
tensely practical. As Andrew Meikle’s apprentice and James Watt’s drafts-
man, Playfair had been a first-hand witness to the work of several great
engineer-entrepreneurs, including not only Meikle and Watt, but also
Matthew Boulton, John Rennie, Josiah Wedgwood, and James Keir. He
had observed the development and success of Boulton’s manufactory
at Soho, the world’s first factory to be organized and run in ways that
we would recognize today. Playfair’s first publication on economics ap-
peared in 1785, but it contained no charts. A preliminary edition of the
Atlas,with engraved charts, also appeared in 1785 – this was privately
circulated to a select few for criticism. The Commercial and Political Atlas
of 1786 was the first publicly available volume to contain charts, and it
exhibits 43 variants of the time series line graph together with a solitary
bar chart. Playfair issued a second edition, which was little changed, in
1787. Despite isolated critical approval, this foray into publishing made
neither riches nor reputation for Playfair and he left England in 1787 to
seek his fortune in Paris. British industry and commerce were leading the
world, and Playfair believed that with his experience at Boulton & Watt
he would be well placed to profit in a France striving to industrialize and
catch up to her neighbor and traditional enemy.
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Introduction 7
Playfair planned to establish a rolling mill in Paris and, for the moment
at least, it seemed that writing was to take a back seat to engineering.
Although the plan was approved by Louis XVI himself, it appears that
the venture never got off the ground because Playfair was soon involved
in other speculative schemes. One ambitious project was the d´
ebacle
subsequently known as the Scioto speculation. This was a complicated,
fascinating, and murky business originating far beyond the borders of
France, in postrevolutionary America, and although much has been doc-
umented by historians of early American corporations (Belote, 1907),
many of the crucial details remain shrouded, including Playfair’s precise
role in the collapse of the scheme. At the end of the Revolutionary War,
syndicates were formed to purchase large blocks of land and to sell in-
dividual tracts at an advanced price to European settlers. The American
Scioto Land Company established a branch in Paris to peddle the idea to
minor French aristocracy, many of whom were becoming increasingly
uncomfortable in the rapidly changing political climate of 1788–9. Joel
Barlow, the unilingual American representative of the Scioto Land Com-
pany in Paris, needed an English-speaking partner who was familiar with
the language and local customs. Playfair fit the bill. Although large sums
were subscribed and several hundred French citizens emigrated to the
wilds of Ohio, the venture ultimately failed and Playfair was accused of
hastening the collapse by embezzling funds. However, mismanagement
on the part of Barlow and Playfair, coupled with the unpreparedness of
the early settlers in a difficult environment, seem equally likely reasons
for the failure. Scioto was not the only problematic speculation to oc-
cupy Playfair during his years in Paris; he was involved in other legal and
financial entanglements and was forced to leave France shortly before
the Terror of 1793.
He spent the years between 1793 and 1814 in London, with occasional
excursions to the Continent. During this time, he published several books
that included charts, the most notable being Lineal Arithmetic (1798),
the Statistical Breviary (1801), and An Inquiry into the Permanent Causes
of the Decline and Fall of Powerful and Wealthy Nations (1805). In 1809–
11 he published the illustrated British Family Antiquity Illustrative of
the Origin and Progress of the Rank, Honours, and Personal Merit, of the
Nobility of the United Kingdom,which included chronological diagrams;
hopes of substantial subscriptions from the aristocracy were evidently the
motivation behind this mammoth nine-volume endeavor. In business,
always seeking new ways of making money, he attempted to import some
of the freewheeling financial schemes that he had used in Paris, but the
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
8Introduction
Bank of England was even less tolerant than the French authorities and
Playfair narrowly escaped prosecution in 1797. He continued to write –
his output numbering more than a hundred books and pamphlets –
but without great monetary success. His many writings on economics
include a critical edition (1805) of Adam Smith’s An Inquiry into the
Nature and Causes of the Wealth of Nations.Smith’s admirers thought
Playfair’s additions and commentaries to be insufficiently respectful,
and the edition was not well received. In general, his political views
were expressed in typically brash and forthright fashion, and his candor
did little to win him friends. He coedited a daily paper, the Tomahawk,
and also a weekly, Anticipation,but both soon failed. He frequently fell
back on his engineering training, working as a gun-carriage maker, and
from time to time he supplemented his income by dubious means. One
swindle led to conviction at the Court of King’s Bench in 1805.
When the Bourbon monarchy was restored, Playfair returned briefly
to France. In 1789, he had initially been in favor of the revolutionaries,
but their later excesses forced a change of mind and his subsequent roy-
alist views, which he was never shy to express, made him unwelcome
during the years of the republic and empire. In 1814, after the accession
of Louis XVIII, Playfair felt that he might, once again, seek better times
in Paris. He was appointed editor of an English language newspaper,
Galignani’s Messenger,and he wrote a number of pieces on the state of
France. He does not seem to have engaged in business affairs of any conse-
quence, and he eventually fled the country after being convicted of libel.
Once more, his reckless and outspoken opinions had constrained his
options.
Playfair lacked money in his final years. Writing did not produce the
anticipated income and, in worsening health, he lost his enthusiasm
for the grand scheme. Although his two sons were independent by this
date, life was not easy for a man who was supporting a wife and two
daughters, one of whom was blind. In 1816, short of cash, he descended
to attempted extortion when he tried to sell some papers alleged to relate
to the great Douglas Cause of half a century earlier. Lasting seven years,
the Cause had been the longest and most expensive legal proceeding in
Scottish history. The documents that Playfair offered to Lord Douglas
were relevant to the alleged imposture of newborn twins conducted in
Paris many years previously. The papers were said by Playfair to have
cast doubt, yet again, on the legitimacy of the Douglas inheritance. These
papers almost certainly never existed; they were merely a prop in Playfair’s
plan to extort money from one of the richest men in Scotland. Because
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Introduction 9
of Douglas’s resistance and Playfair’s weak evidence, the blackmail did
not succeed and indeed did not come to light until recently (Spence &
Wainer, 1997). This shameful affair demonstrates Playfair’s straitened
financial situation and his readiness to ignore the law when it suited his
purposes.
The last few years of Playfair’s life saw a renewed interest in eco-
nomics, and Playfair’s final publications include some very fine charts
(Playfair’s two letters on agricultural distresses, 1821, 1822). These late
worksexamined the difficulties experienced by English farmers in the
early 19th century. Playfair died on 23 February 1823 in Covent Garden,
likely in the house at No. 43 Bedford Street. He was survived by his wife
and four of his children, one of whom, Andrew William, had emigrated
to Canada where he was prominent in the military and successful in pri-
vate business, eventually founding the town of Playfairville not far from
the capital, Ottawa. Andrew William persuaded his older brother John to
join him, and their descendants have prospered and spread throughout
Canada.
During his life, despite the interest and approval of a select few, William
Playfair’s invention of statistical graphs went largely unacknowledged.
Although he was a tireless advocate for his charts, he made few converts.
Hisobituaries ignored the graphical inventions and concentrated on his
political and economic writings, which were not held in high regard by
his contemporaries, although they have attracted renewed interest today.
One apologist wrote:
HadMr. Playfair cultivated his mechanical genius, there is no
doubt, that he would not only have obtained considerable emi-
nence, but have rendered no inconsiderable service to this country.
Unhappily, however, for his own interests, he had the ambition to
become an author. (Author unknown, Edinburgh Annual Register,
1823, 332)
18TH-CENTURY BARRIERS TO STATISTICAL CHARTS
To day, most people think that statistical graphs are such simple and ob-
vious creations that almost anyone could have invented and published
them. Indeed, it is their simplicity that accounts for much of their appeal,
and that is why we give scarce thought to the ingenuity required to invent
and promulgate statistical charts. Familiarity has dulled our apprecia-
tion of their significance, diminishing the importance of their creator,
whose name until recently was largely unknown, even to professional
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
10 Introduction
statisticians. But the idea of devising and publishing statistical graphs
was not obvious two centuries ago and, even today, the form is not nearly
so naive and self-evident as it might first appear (Cleveland, 1985; Tufte,
1983; Kosslyn, 1994; Spence & Lewandowsky, 1990; Wainer & Velleman,
2001; Wainer, 2000, 2005).
Large collections of economic statistics were widely available – and
had been since the time of Graunt and Petty – more than a century be-
fore Playfair thought of publishing such data in pictorial form. The data
necessary for the invention of statistical graphs were present in abun-
dance, but no one else had the inspiration to represent them as pictures.
There were various impediments to the publication of illustrations in se-
rious writing. There were philosophical objections, concerns regarding
accuracy and misrepresentation, and technical barriers to publication.
Plants have been portrayed in print since the introduction of the print-
ing press in the 15th century. From early Renaissance herbals, through
pictures of Baroque gardens, to increasingly naturalistic depictions of
plants and flowers in the 17th and 18th centuries, printed illustrations
of natural history had become fairly common and accepted. But there
is ample evidence to suggest that similar illustration in serious scien-
tific writing was viewed with suspicion, and eminent experimenters like
Robert Hooke, who used illustration, did so with misgivings. About his
Micrographia of 1665, which contained many illustrations, Hooke wrote
that
Pictures of things which only serve for ornament or Pleasure, or
the Explication of such things as can better be describ’d by words
is rather noxious than useful, and serves to divert and disturb the
Mind, and sways it with a kind of Partiality or Respect. (64)
Hooke worried about the possibility of misrepresentation and took great
pains to assure the reader of accuracy, or to point out possible distortions
that the illustrations might produce in the reader’s mind. Tilling (1975)
has pointed out that in the 17th century information in charts produced
by automatic graphical recording devices, such as weather clocks, was
often translated into tabular form and that, with one exception in 1724,
there was no publication of similar charts until the 19th century. Pre-
sumably no value was seen in graphical presentation or, more simply, the
continuous graphical record was not regarded as being as trustworthy
or informative as the corresponding sequence of numbers.
Biderman (1990) and Valois (2000) have argued that a mistrust of
sense perception on the part of Descartes and his disciples was a powerful
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Introduction 11
impediment to the development of empirical methods of investigation,
including the development and use of statistical graphs. This impedi-
ment was to disappear as British empiricist philosophy evolved during
the late 18th century. Beginning with Locke, Berkeley, and Hume, British
empiricists in the 17th and 18th centuries had argued that knowledge
comes from experience whereas the rationalists, such as Descartes, had
maintained that knowledge may be derived solely through reason based
on innate ideas. The empiricists rejected the notion of innate ideas and
argued that almost all knowledge is based on sensory experience. The
Scottish realist philosophers, such as Thomas Reid and Dugald Stewart,
took this idea even further by emphasizing the process of inductive rea-
soning from sense data. In a distinction that still has currency today,
Reid (1764) characterized the difference between rationalist philosophy
and empiricism in two ways. Firstly, rationalists believed that concepts
are known intuitively through reason, as opposed to experience. And
secondly, rationalists maintained that truths could be deduced from
innate ideas, in much the same way that theorems are deduced from
axioms. Mathematical proof was the model for obtaining knowledge.
Although the Scottish empiricists also used deductive reasoning when
appropriate, they attached much greater importance to the inductive
method. Both Thomas Reid and Dugald Stewart, Reid’s intellectual suc-
cessor, rejected much of Cartesian philosophy, preferring to rely on ob-
servation and inductive reasoning. Although more than two centuries
old, their approach is surprisingly modern. Nowadays, we take for
granted that empiricism is at the core of modern scientific method, which
considers that theory should be tested by observation rather than by
intuition.
The Playfair brothers were well acquainted with the leading empiri-
cist philosophers of the Scottish Enlightenment, such as Hume, Reid,
and Stewart, and their own thinking was in the same empirical mold.
John Playfair and Dugald Stewart were firm friends and colleagues and
their approaches to mathematics and natural philosophy were highly
compatible. Indeed, in 1785, John Playfair succeeded Stewart as profes-
sor of mathematics when the latter relinquished his chair to take Adam
Ferguson’s chair in moral philosophy. As a boy, William Playfair had ab-
sorbed the prevailing Scottish empirical approach in science, learning to
represent physical data by line graphs under the instruction of his brother
John (Playfair, 1805, xvi). However, although such charting had become
common among natural philosophers for tracking their private exper-
imental data, they did not employ these devices to buttress arguments
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
12 Introduction
via publication. John Playfair, for example, never made use of the line
graph in his writings (J. Playfair, 1822). It was William Playfair’s genius
not only to apply the line graph to economic data but to see the value in
publication. However, despite the shift to empiricism, a general mistrust
of pictorial representation persisted, with 18th- and early 19th-century
academics reluctant to publish graphs of physical or statistical data, pos-
sibly because of lingering concerns regarding accuracy, or simply the
sheer technical difficulty of publication.
Copperplate Engraving
Asignificant barrier was the process of printing illustrations. As
’Espinasse (1962) has noted, in the 17th century, scientists were ver-
satile, used to manufacturing their own apparatus and to dealing with
tradesmen and craftsmen, whereas by the 18th century, scientists had
become more specialized and were less likely to possess the practical
expertise to engrave their own plates. But William Playfair did have the
skills. He was no academic insulated from the real world; his training
had prepared him perfectly for the invention and production of statistical
charts. Mastering the printing process was easy for Playfair, the engineer
and draftsman, and we know that he even frequently engraved the lines
on the copper plates himself, leaving the more delicate work of lettering
and decoration to the printer.
In the 18th century, intaglio copperplate prints of maps or natural his-
tory were usually signed by the artist and the engraver. The artist’s name
often appeared in the white space beneath the image on the left-hand
side, followed by “del.,” an abbreviation for the Latin delineavit, and
the engraver/etcher’s name would appear on the right side, with “sc.,”
an abbreviation for sculpsit.Inmost of Playfair’s plates one engraver’s
name, Neele, appears in the traditional position on the lower right, but
the early plates by J. Ainslie are signed in the lower left. Playfair’s first en-
graver, John Ainslie (1745–1828) was also an outstanding cartographer,
surveyor, and publisher. Born in Jedburgh in the Scottish Borders, he
is best known for “Ainslie’s Travelling Map” (1783) and his large-scale
map of Scotland (1789). Nine of the first 10 plates in the first edition of
the Atlas are by Ainslie. One of the plates is unattributed and, judging
by the small differences in style from Ainslie’s plates, it may have been
completed by Playfair himself. The first 10 plates are dated 1785 and may
have been the only plates to figure in the preliminary edition that was
privately circulated to Watt, Boulton, and a few others. All except one
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Introduction 13
of the remaining plates, dated 1786, are by Samuel John Neele (1758–
1824), a London engraver and copperplate printer who specialized in
book illustrations and maps. Neele signed most of his engravings using
only his surname and the address of his business at 352, Strand, London.
Both engravers were craftsmen of great ability, and many fine examples
of their maps still survive. However, the quality of engraving in Playfair’s
charts is not comparable to the excellence of the cartography produced
by these two men in their customary work. The likely explanation is that
Playfair required his work to be done economically and requested the
craftsmen to minimize expensive frills.
Acopperplate engraving is an image taken from an engraved cop-
per plate. Although copperplate printing developed as early as the
14th century, it was only during the 17th and 18th centuries that cop-
perplate engravings were widely used for illustrated works in France and
England. Copperplate remained the standard until the early 19th century
when engraving on more durable steel plates took over. A plate of bright,
burnished copper is first coated with a ground, usually a hard wax, and
then the desired image is traced with a needle. The ground is then re-
moved. Guided by the traced lines, the engraver uses a burin, a metal
tool with a sharp point, to engrave onto the copper plate. Metal shavings
are cut away by the burin. These shavings, or “burrs,” must be detached
bya“scraper,” another cutting tool. The deeper the cut, the stronger
the printed lines will be. The plate is then warmed, inked, and passed
through a press with the sheet of paper to be printed.
Etching was also used to produce the design, which was first drawn
with a needle to penetrate an acid-resistant wax coating previously ap-
plied to the plate. The plate was then immersed in a bath of acid, which
etched those areas exposed by the needle. Deeper lines were produced
by re-immersion in the acid, after lines that were not to be deepened had
been stopped out with acid-resistant varnish. Without a microscope it is
difficult to distinguish engraving from etching, with the latter producing
slightly fuzzier lines. Most illustrators in the 18th century employed a
combination of the two processes. Neele and Ainslie probably used both
methods to create the decoration, framing, titles, and other lettering, but
in many cases Playfair probably engraved the data lines himself. Several
plates show evidence of careless or inexpert use of the burin to engrave
the data lines and the grid lines; in the third edition, reproduced here,
there are obvious blunders in a few plates and minor errors in several
others.
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
14 Introduction
Because copper is a soft metal, it was rare to take more than
1000 impressions before the plate deteriorated. This is one of the rea-
sons why many works at the time appeared in several editions soon after
the original. These new editions provided an opportunity to add new
information, as Playfair did with the third edition of the Atlas.
THE COMMERCIAL AND POLITICAL ATLAS (1786–1801)
In 1785, Playfair first wrote about economics in The Increase of Manufac-
tures, Commerce, and Finance, with the Extension of Civil Liberty, Proposed
in Regulations for the Interest of Money; this extended essay contained no
charts. In the same year, he also circulated a version of the Commercial
and Political Atlas to selected acquaintances. He sought criticism that
would help him to improve the published version. He sent copies to
both Watt and Boulton, writing to the latter:
Ihave taken the liberty of asking if you will do me the favour to
accept of the first number of a work which I hope may be one day of
considerable utility. I must beg leave as a particular favour that you
will oblige me with your remarks and any hints for amendment or
improvement that may occur to you as I wish to make it as perfect
&complete a work as I am able. (20 September 1785)
A similar letter must have been sent to Watt who replied:
I can think of nothing in addition to your plan, except that it
might be proper to give in letter press the Tables from which the
Charts have been constructed ...for the charts now seem to rest on
your own authority, and it will naturally be enquired from whence
you have derived your intelligence. A general chart showing the
increase & decline of various articles of exportation & importation,
or particular charts of the same would also be useful, but my mind
is so much turned to other pursuits that I cannot pretend to direct
you. (10 October 1785)
Watt, ever the careful scientist and engineer, was concerned with the
accuracy and the provenance of the data and Playfair, who revered Watt,
was persuaded to include tables in the first (1786) and second (1787)
editions of the Atlas.By1801, Playfair had decided to ignore Watt’s
advice, and no tables are to be found in the third edition.
The bulk of the Atlas examined English commerce with other nations
during the 18th century. The first edition was published in foolscap
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Introduction 15
folio, cut in the size of 8 ×13 inches (216 ×330 mm), with the type
set in landscape format rather than in the more common portrait for-
mat. The charts appear on separate pages at the same size as the text.
From about the 16th to the 18th centuries, illustrations were printed
using engraved copper plates, inked and wiped so that the ink remained
only in the incisions. However, this intaglio method was at odds with the
printing of text where the type was in relief. Consequently, the cheap-
est solution was to print the illustrations separately (Biderman, 1981;
Spence, 2000). Playfair was probably driven to the less common land-
scape format to accommodate charts that would have been much less
successful if they had been constrained to a vertical format. In using
the more effective landscape layout, Playfair seems to have anticipated
modern ideas regarding the optimal aspect ratios of graphs (Cleveland,
1985). Although the second edition uses an identical layout, by the third
edition, Playfair had adopted a more conventional vertical format for
the text. Partly because of this change, some charts, which are two to
three times the page size, appear as folded flyouts from the main body
of the work. Playfair could have achieved the same end by printing these
charts sideways – and he did this in most instances – but he clearly
wanted to be able to present the more important charts with sufficient
detail.
In the first edition, 43 of the 44 charts plotted pounds sterling on the
ordinate against time on the abscissa. In addition, a solitary bar chart –
an oddity made necessary because Playfair did not have sufficient data
to construct a line graph – was the only chart that did not include time
as a dimension. The bar chart was directly inspired by Priestley’s (1765)
chronological charts and Playfair must have been aware of the irony when
he apologized for the anomaly: “This Chart ...does not comprehend any
portion of time, and it is much inferior in utility to those that do” (Atlas,
1786, 101).
The graphs in the three editions of the Atlas were remarkably similar
to those in use today; hachure, shading, color coding, and grids with
major and minor divisions were all introduced in the various editions
of the Atlas.Actual, missing, and hypothetical data were portrayed, and
the kind of line used, solid or broken, differentiated the various forms.
Playfair filled the areas between curves in most of the charts to indicate
accumulated or total amounts. All included a descriptive title either
outside the frame (as in the first edition) or in an oval in the body of
the chart (as in the third edition). The axes were labeled and numbered
where the major gridlines intersected the frame.
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
16 Introduction
Differences between the Editions
There are few differences between the first and second editions. The first
edition contains 44 charts in 40 plates – copperplates colored by hand
in three or four colors. Twenty of these charts represent trade between
England and other countries. The line of imports is stained orange-
yellow and that of exports is stained red. The area between the curves is
abluish-green when the balance is favorable to England and pink other-
wise. In the second edition, the charts differ mainly in the loss of color.
Before the invention of multiple-layer color printing, printed black ink
intaglio illustrations were sometimes colored freehand. Water color was
applied to the image to enhance the design in much the same fashion as
in a watercolor painting. By the 18th century copper engravings colored
by hand were fairly common. Colorists were employed by the artist or
publisher to produce the finished color prints although, in Playfair’s case,
there is reason to believe that he colored many of the charts himself to
save money. To color or not to color forced a choice on the engraver. An
engraving that was not to be colored had to represent color and shape by
means of the engraving alone. Hues were represented by cross hatching
or stippling with dots. Outlines were generally engraved with a heavier
hand, resulting in a print with the look of a pen and ink drawing. On
the other hand, copperplate engravings that were to be hand-colored
depended on the colorist to add detail and shape. Engraved lines were
kept light so as to not interfere with the water colors. The heyday of hand-
coloring lasted for about a hundred years, from the middle of the 18th
to the middle of the 19th centuries, when new color printing processes
became available.
Playfair moved to France in 1787 and probably had neither the time
nor the money to do the laborious hand-staining that was necessary to
color the charts. He replaced the two major colors, representing favor-
able and unfavorable balances, by using hachure (for blue-green) and
stippling (for pink). Otherwise, essentially the same charts appear in
both editions and also in Lineal Arithmetic (1798), where color, hachure,
and stippling are all used, perhaps an indication of slightly easier fiscal
circumstances for Playfair.
There are much more substantial differences in the third edition.
Instead of 40 plates containing 44 charts, there are 28 plates containing
33 charts. The most significant omission is the bar chart showing the
exports and imports of Scotland. Gone also are charts showing trade
data for England, Holland, and the United States. Three charts showing
aspects of the national debt as it related to annuities and interest rates
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Introduction 17
were also dropped. Missing also are the five charts, attributed to James
Corry, representing economic data from Ireland. Two new charts were
added. The first, in Plate 19, is a rather elaborate large area chart on a
flyout showing the annual revenues of England and France as well as the
interest on debt. This chart includes a chronological display at the top
which shows the reigns of English, British, and French monarchs. The
other new plate is not numbered, although it is given a figure of 26 in
the index and referred to as Chart XXVI in the text.
Although much of the data are brought up to date, the tables of num-
bers are no longer incorporated. James Watt had advised their inclusion
in the earlier editions to allay possible concerns regarding provenance
or accuracy. In fact, as we shall see, the tables call attention to Playfair’s
lack of concern for accuracy, and so Watt’s wise counsel had an effect
opposite to what he intended. Playfair’s goals were didactic and at times
polemical, rather than analytical, and his freehand drawing of the vari-
ations in imports or exports is sometimes hard to reconcile with the
numbers. While he certainly made small errors and technical mistakes,
the most egregious problems concern his interpolations between data
points. On many occasions, the ups and downs of the lines are fanciful
and probably reflect Playfair’s prejudices rather than the likely values of
the missing data.
Playfair did not believe that the accuracy of a graph could exceed that
in a table. Nor did he feel that such accuracy was necessary:
The advantage proposed by this method, is not that of giving a
more accurate statement than by figures, but it is to give a more
simple and permanent idea of the gradual progress and compara-
tive amounts, at different periods, by presenting to the eye a figure,
the proportions of which correspond with the amount of the sums
intended to be expressed. (Atlas, 1801, ix–x)
Nevertheless, it is surprising that he was not more concerned with ac-
curacy since he was well aware that his new methods might arouse sus-
picion. He tackles the issue head-on in the advertisement that opens the
first edition:
As to the propriety and justness of representing sums of money,
and time, by parts of space, tho’ very readily agreed to by most
men, yet a few seem to apprehend that there may possibly be some
deception in it, of which they are not aware...(Atlas, 1786, iii)
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
18 Introduction
As we have noted, James Watt also had stressed the importance of accu-
racy, and he persuaded Playfair to include at least some of the original
data in tabular form. But, alas, despite Watt’s counsel, the Atlas is not a
model of precision. Several arithmetical errors and careless drawing are
evidence of rushed production. Some of the curves that connect the data
points seem to have derived their shapes from Playfair’s opinion of how
the intervening data should look. His curves are drawn freehand, often
somewhat crudely, betraying a lack of practice in the demanding skills
of engraving. We give some examples of these peculiarities and technical
flaws:
rArithmetic.Inthe section labeled “Contents of the plates in num-
bers” (first ed., 1786, 21–2) there are minor, but careless, arithmeti-
cal errors in the tables. Several of the “balances” fail to agree with
the difference between “exports” and “imports.” These faults, in
themselves, are trivial, but such carelessness throws into doubt the
accuracy of the remaining numbers in these and the other tables.
rPlate 4.This plate, which has the same number in all editions of
the Atlas,illustrates several problems. While Playfair’s rendering
of the raw data (Table IV in the first edition) is reasonably faith-
ful in the first and second editions, there are large discrepancies
between the numbers and the curves in the third edition. In our
Figure 1, we have superimposed smoothed blue (import) and green
(export) curves that are consistent with the tabulated data. There
are substantial differences between our curves and Playfair’s depic-
tion, which is also considerably different from the graph that he had
published 15 years before. A possible explanation is that Playfair had
discovered more accurate data in the interim – if so, he does not
allude to this in the text. Since there is little doubt that scholars
in the 18th century were concerned about the misrepresentations
that pictorial methods might bring with them, Playfair may have
damaged his case for adoption by constructing charts that did not
faithfully reflect the numbers.
Another curiosity in Plate 4 is that the engraved line for im-
ports terminates around 1785 and, similarly, the line of exports
stops at about 1792, although the hand-staining of both lines con-
tinues until 1800. There is no obvious reason for these discrepan-
cies, which do not occur in other charts where, if the engraved line
terminated early, the color stain usually did not continue beyond
that point.
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Figure 1 Plate 4 (1st, 2nd, and 3rd eds.) showing trade with the
West Indies. We have matched the scales of the plates to allow a
direct comparison of the curves. This has the consequence of
producing a slightly distorted third edition plate that is wider
than the original. The central panel was constructed using the
data from Table IV in the first edition and is discrepant from the
curves in the third edition. Note that our reproduction of the first
edition plate is taken from microfilm and lacks the color of the
original.
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
20 Introduction
rPlate 17 (Plate 22 in the first edition).This plate, although engraved
as number 17, is indexed as Plate 16 in the third edition. The Uni-
versity of Pennsylvania’s plate number has been altered by an un-
known hand to conform to the index, whereas, as our Figure 2
shows, the University of Toronto plate number remains unaltered.
Minor mixups such as this show that Playfair, his engravers, and
printers worked at speed, with small errors creeping in. Even today,
errors of this sort are not unknown in publishing. However, a more
interesting aspect of this chart is Playfair’s rendering of the tabulated
data into curves. Although his representation does not reflect the
actual numbers with scrupulous accuracy, this is a minor problem.
It is his conception of how the data have varied during each suc-
cessive ten-year interval that is curious. As far as we know, Playfair
did not attempt to fit the data for each individual year in the several
decades – even if these data had been available – and therefore his
curve is likely an interpolation. Our conservative smooth interpola-
tion yields just three points where the line of imports crosses that of
exports, whereas Playfair’s graph exhibits seven such intersections.
Thus one gains the impression that the balance of trade oscillated
between England and the Channel Islands during each decade of
the 18th century, when, in fact, there was probably only one period,
from about 1710 until about 1740, during which the balance of trade
favored Jersey, Guernsey, and Alderney. There are other plates that
exhibit variations in the curves that are not easy to reconcile with
the data given in the tables from the first two editions.
rPlate 5.Different copies of a work from the same printing show dif-
ferences. These variations highlight the handmade nature of books
two centuries ago. An example is shown in our Figure 3, which re-
produces Plate 5 from the third edition of the Atlas.The upper plate
comes from the copy in the Annenberg Rare Book and Manuscript
Library at the University of Pennsylvania and the lower plate comes
from the volume in the Thomas Fisher Rare Book Library at the
University of Toronto. There are noticeable differences in the hand-
staining of the two charts. After two centuries, it is difficult to know
whether differences in color are a result of differential aging or
whether such variation was present in the originals. It seems that
the upper version lacked original staining in the frame and title oval.
The upper plate has been more hastily colored and the areas between
the lines of import and export are much less carefully covered than
in the lower plate.
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Introduction 21
Figure 2 Plate 17 (3rd ed.) and Plate 22 (1st and 2nd eds.) showing trade
with Jersey, Guernsey, and Alderney. This plate is numbered 16 in the index
to the third edition. We have matched the scales of the plates to allow a
direct comparison of the curves. The central panel was constructed using
the data from Table IV in the first edition. Playfair’s curves exhibit seven
crossings, whereas our curves cross only three times. Note that our
reproduction of the first edition plate is taken from microfilm and lacks
the color of the original.
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
22 Introduction
Figure 3 Plate 5 (3rd ed.) showing trade with North America. These two
versions of the same plate are reproduced from different copies of the book.
Differences in quality and quantity of the hand-staining are noticeable. A
common engraving error in the line of imports is seen at the extreme right, with
the hand-stained yellow-orange line indicating the correct branch.
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Introduction 23
There is an engraving error, just before 1800, where the engraved
line of imports takes two paths. Playfair probably inscribed the lower
branch in error and subsequently added the correct path. The fault
is scarcely noticeable since the correct curve is clearly emphasized
by the yellow-orange staining.
The engraver’s name, Neele, is at the lower right in the University
of Toronto’s copy but not in the University of Pennsylvania’s copy,
and there are eight other similar instances. There are also discrepan-
cies in plate numbering for three of the plates and also between the
plates and the index. These anomalies suggest that the plate num-
bering and engraver attributions may not have been engraved on
the original copper plate but were added later.
Graphical Innovation
Despite the minor numerical errors, the technical slips, and graphical
functions that are occasionally more fanciful than accurate, all three
editions of the Atlas introduced an astonishing number of novel charting
constructions that are still in common use today (Tufte, 1983; Biderman,
1990; Costigan-Eaves & Macdonald-Ross, 1990). We shall comment on
only a small sample of Playfair’s inventions (some of which are seen in
the first and second editions only), and we invite the reader to discover
others:
rThe time series line graph.The third edition has one simple time
series line graph (Plate 26). Although all other charts use a line to
display variation in amounts, most also include staining or shading
between the curves and may be described as area charts. Playfair
introduced several different variations of this form.
rThe divided surface area chart.Agood example is found in the first
edition (Plate 38, after p.153). This chart does not appear in the third
edition, and is one of several charts that are due to James Corry, after
the fashion of Playfair. In the third edition, Plate 19 is a divided area
chart. Area charts are ideal for showing trends when the variation
in two or more time series must be shown simultaneously. The area
between each line and the abscissa is filled with a color for that data
series, with the colors in the lower areas occluding those higher up.
The use of color (or hachure or stippling) serves to emphasize the
ways in which accumulated amounts have varied.
rThe bar chart.This chart, inspired by chronological diagrams
(Funkhouser, 1937; Wainer & Spence, 2005), was introduced in the
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
24 Introduction
first edition (see our Figure 4), but by the third edition this chart
had disappeared and Playfair did not use the form again.
rTitles and textual descriptions.The first edition used descriptive titles
above the chart, outside the frame. Explanatory notes regarding the
scaling of axes appeared below the frame. Curves and stained areas
in the charts were labeled. Other information often also appeared
below the frame, for example, the engraver’s name, or the date. By
the third edition, Playfair had arrived at a more designed – and
more expensive – look. The title was relocated to an oval or other
shape within the frame; decoration of the text in this caption was
common, as in Plates 19 and 21. Our Figures 1 and 2 show this
evolution.
rFraming.Thecharts were invariably framed. In the first edition,
this consisted of a simple double-lined box, similar to the ones that
we have used in our illustrations in this Introduction. By the third
edition, the frame included a stained border just inside the double-
lined box. This provided a space for the labels and scale values, and
it made the chart more pleasing to the eye. Our Figures 1 and 2
provide an illustration of this evolution.
rColor coding.Plate 17 provides a good example. A thick red hand-
painted line is used for Exports; a yellow-orange line for imports;
solid blue-green fill color between the export and import lines shows
when the balance of trade is positive – exports exceed imports – and a
pink-red solid fill shows when the balance of trade is negative. Thus
color is used by Playfair to emphasize the qualitative differences
between the time series and the quality and quantity of the varying
accumulated amounts.
rHachure and stippled dots.Where color was not available, as in the
second edition, Playfair adopted the engraver’s practice of simulat-
ing dark colors by hachure and lighter colors by stippling.
rLabeling of axes.West border: vertical label “Money”; north border:
horizontal label “Years”; south border: time scale; east border:
money in pounds sterling. This scheme forces the graph to have
aframe, unlike common practice today where the labels and scales
tend to be on the west and south sides only.
rGridlines.Major gridlines are engraved more heavily than minor
gridlines. Plate 1 is a good example. Minor vertical gridlines are
not used where there are no data (as indicated in the tables of the
first edition). Presumably, after 1780, Playfair had data for each
individual year.
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Figure 4 Plate 25 (1st ed.) showing the trade of Scotland during 1781.
25
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
26 Introduction
rSuppression of nonsignificant digits.Plate 3 is a good example. The
scale is implicitly defined by the first label, where full precision is
used, thus indicating the value of the intervals between gridlines
(however, Playfair is not consistent in his use of this device).
rTime period indicators.InPlate 19 of the third edition, the reigns of
the kings of England, Britain, and France are shown at the top in
the style of the chronological diagram that Priestley used in 1765 to
show the life spans of significant personages from classical antiquity.
Another example is found in Plate 31 of the first edition (after p.133);
the upper three black horizontal bars indicate times of war. In the
corresponding Plate 25 in the third edition, Playfair has dropped
these bars.
rEvent markers.InPlates 6, 20, and 21, Playfair uses vertically ori-
ented text in the body of the figure, positioned to mark important
historical events that may have had some influence on the subse-
quent trend in the time series.
rTheoretical/hypothetical/projected values.Plate 21 shows the pro-
jected reduction in the national debt after a “Sinking Fund” was
established by the Prime Minister, Pitt the Younger. The national
debt had assumed staggering proportions due to the rebellion of the
American colonies, and William Pitt planned to retire the debt by
imposing new taxes. But taxation alone was insufficient. In 1786,
Pitt, who became Prime Minister in 1783, introduced a Sinking
Fund, using an idea first implemented by Walpole in the 1720s.
Each year, £1,000,000 of the surplus revenue raised by new taxes
was to be added to the fund. The accumulated interest was to be
used to pay off the national debt. The system was extended in 1792
so as to take into account any new loans taken by the government.
The system worked in peacetime – with regular annual surpluses –
but, after the country went to war in 1793, the debt was redeemed
by new borrowing at higher rates of interest. The chart in Plate 21
shows Playfair’s model for the time course of the reduction in the
national debt. Playfair does not explain the mathematical basis for
the construction of his curves, other than to allude to compound
interest, but he is skeptical of the claims of others, whom he refers
to as “arithmetical pedants.”
Although the national debt was a growing concern in 1786,
and Playfair discusses both the debt and the Sinking Fund (1st
ed., 131–3), his earlier graphical approach was rather different in
Plate 30 in the first edition (before p.131). Plate 21 (3rd ed.) and its
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Introduction 27
often-reproduced empirical partner, Plate 20 (3rd ed.), do not ap-
pear in the first or second edition.
rSolid and broken lines.While most of the lines and cur ves that Playfair
drew are solid, he occasionally used broken lines. Sometimes he
seems to have intended this to indicate uncertainty in the data. For
example, in Plate 33 of the first edition, he uses a broken line as a
companion to a solid line representing the national debt to indicate
that “the state of the debt is exact, or very nearly so.” Thus the broken
line indicates a lower degree of confidence in the accuracy of the
data. In the Statistical Breviary,broken lines are used to indicate the
relationship between population size and taxation (Charts 1 and 2).
In this instance there is no suggestion that the data are inaccurate.
Abroken line is also used with the outer of two concentric circles
that represent the area of France and the other territories that it
controlled (Chart 2). It is difficult to know whether Playfair used
the broken line to indicate a lack of confidence in the data in this
case.
THE STATISTICAL BREVIARY (1801)
In the Statistical Breviary,Playfair presented statistical data for European
countries at the beginning of the 19th century. He used charts since he
believed that “making an appeal to the eye when proportion and mag-
nitude are concerned, is the best and readiest method of conveying a
distinct idea” (4). The most important graphical innovation in this
volume was the pie chart. The intellectual origins of the pie chart re-
main obscure – although Playfair acknowledged and wrote about the
inspirations for the time series line chart and the bar chart, he was
silent regarding the pie chart. This diagram seems almost certain to
have derived its inspiration from the logic diagrams of Leibniz and Euler
(Spence, 2005), but it is likely that pies, circles, and intersecting circles
were such simple and familiar forms to Playfair that he did not think
explanation or comment was necessary. Nevertheless, his use of pie and
circle diagrams, which had until then only been used to illustrate con-
cepts in mathematical logic, was revolutionary. Playfair was a capable
and inventive adapter of ideas from other domains (Spence & Wainer,
2001), and his adaptation of logic diagrams to portray and compare
empirical data was ingenious.
The first chart in the Statistical Breviary depicts European countries
before the French Revolution of 1789, and the second chart shows how
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
28 Introduction
circumstances had changed by 1801. Circles represent the land areas; for
example, Russia, the largest country, is symbolized by the circle of great-
est diameter, while small nations like Portugal call for tiny circles. Just
below the horizontal diameter of most circles, Playfair has inscribed the
values of the areas in square miles. The charts also depict the sizes of the
populations and the revenues of the countries, and whether individual
countries were maritime powers (area stained green) or nonmaritime
powers (stained red). The sizes of the populations and the tax revenues
are represented by vertical red lines on the left of each circle and by the
vertical yellow lines on the right. The dotted lines joining the tops of the
lines show the tax burden on the populations. However, as Funkhouser
(1937) noted, “the slope of the line is obviously dependent on the diam-
eter of the circle” and so it cannot serve as an accurate index of the tax
burden. Playfair probably intended the reader simply to note whether
the slope was positive or negative.
To show how some countries were subdivided, Playfair used several
strategies. For example, the Russian empire was divided into European
and Asiatic dominions with the former represented by the inner circle
and the latter by the surrounding annulus. This diagram uses the two
distinct areas to represent the distribution of the empire between the
two continents. The Asiatic dominions were represented by the annu-
lus, which was stained green indicating a sea power. The inner circle
was stained red to indicate that the European dominions were land
powers. The Turkish empire was harder to accommodate since it was
spread across three continents: Asia, Europe, and Africa. Three concen-
tric circles would have made visual comparison of the areas even more
difficult than in the case of the Russian empire. Playfair’s whole pur-
pose in creating these diagrams that represented quantity by area was
to make comparisons effortless and the data more memorable. Playfair,
who appears to have recognized and understood the perceptual issues
involved, realized that concentric circles yield areas whose sizes are hard
to compare accurately. Accordingly, he divided the circle representing the
Tur kish empire into three sectors proportional to the Asiatic, European,
and African land areas. Again, he used colored stains: green to sig-
nify maritime power (the Asian sector); red to denote land power (the
European sector); and yellow (the remaining African sector). Playfair
gave no rationale for his use of these particular colors, but this diagram
was the first pie chart to display empirical proportions and to distinguish
the component fractions by color.
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Introduction 29
Playfair introduced three new statistical diagrams in the Breviary:
the circle chart, the pie chart, and a figure to show joint properties,
similar to a Venn diagram. Like the line and bar charts, introduced 15
years earlier, his designs have still not been materially improved upon
in 2005.
PLAYFAIR AND THE PSYCHOLOGY OF GRAPHS
During the past two decades, cognitive science has played an important
role in advancing our understanding of the power and utility of statis-
tical graphs. Graphs achieve their success by capitalizing on the basic
perceptual and cognitive capacities of human beings. However, inter-
est in the psychological aspects of charts is not new. Indeed, Playfair
seems to have well understood that our cognitive and perceptual capaci-
ties were critically important (Costigan-Eaves & Macdonald-Ross, 1990).
Perhaps we should not find this surprising since the Playfair brothers were
well acquainted with the ideas and methods of the Scottish empiricist
philosophers – in particular Hume, Reid, and Stewart – whose enquiries
focused on questions in perception and cognition that continue to oc-
cupy experimental psychologists.
Playfair believed that graphs would be a powerful aid to memory;
intuitively, he appreciated that visual memory was more robust than
memory for words or numbers. When he was searching for a better way
of presenting tabular data he said (Atlas,1801, xiv), “a man who has
carefully investigated a printed table, finds, when done, that he has only
avery faint and partial idea of what he has read.” He also appreciated that
intuitive visual comparisons of size could be made much more rapidly,
and almost as accurately, than by mental arithmetic computations, using
the numbers themselves. He claimed that
The advantages proposed by [the graphical] mode of represen-
tation, are to facilitate the attainment of information, and aid
the memory in retaining it: which two points form the principal
business in what we call learning. ...Of all the senses, the eye gives
the liveliest and most accurate idea of whatever is susceptible of
being represented to it; and when proportion between different
quantities is the object, then the eye has an incalculable superiority.
(Breviary, 1801, 14)
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
30 Introduction
He understood that
the eye is the best judge of proportion, being able to estimate it with
more quickness and accuracy than any other of our organs ...this
mode of representation ...givesasimple, accurate, and permanent
idea, by giving form and shape to a number of separate ideas, which
are otherwise abstract and unconnected. (Atlas, 1801, x)
He anticipated modern ideas in cognitive psychology such as depth
of processing by noting that people remember information better when
they process it in a meaningful rather than superficial way: “Informa-
tion that is imperfectly acquired, isgenerally as imperfectly retained”
(Atlas,1786, 3). He speculated that his charts were an aid to meaningful
memorization:
Whatever presents itself quickly and clearly to the mind, sets it
to work,toreason, and think; whereas it often happens, that in
learning a number of detached facts, the mind is merely passive, and
makes no effort further than an attempt to retain such knowledge.
(Breviary, 1801, 6–7)
Playfair’s circle and pie diagrams were intended to facilitate the com-
parison of land areas; comparing the irregular shapes formed by the
boundaries of countries in a conventional atlas was problematic and or-
dering countries by size was a difficult visual task. Playfair’s solution was
to use a common shape and thus exploit the eye’s capability of making
comparative judgments with high accuracy; “for where the forms are not
similar, the eye cannot compare them easily nor accurately” (Breviary,
1801, 15). Playfair was able to offer remarkable insights into the cogni-
tive science of graphical perception two centuries before the flowering
of modern cognitive neuroscience.
PLAYFAIR’S LEGACY
Beniger & Robyn (1978) observed that, beginning with maps of North-
ern Mesopotamia, there was a 3,000-year-old tradition of representing
physical space (the world) by space (the map). Although sufficiently
inspired by mapmakers to use the word “atlas” in the title of his trea-
tise, Playfair ended their monopoly on the use of spatial displays. His
genius was to realize that nonspatial quantities such as expenditures and
historical time could be represented by physical space and that such
representation offered advantages denied to tabular presentation. But
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Introduction 31
others did not share his conviction that he had found a superior way of
presenting data, especially in his own country where concerns regarding
accuracy were heightened by Playfair’s carelessness, brashness, and dis-
reputable personal reputation (Funkhouser & Walker, 1935; Funkhouser,
1937; Spence & Wainer, 1997). He was received more kindly on the con-
tinent where Humboldt thought highly of his creations (see Hankins,
1999), but there was still considerable opposition from many statisti-
cians. Adoption of the new methods had to wait until the second half
of the 19th century when Minard and Bertillon used some of Playfair’s
inventions in their cartographical work (Palsky, 1996; Friendly, 2002).
In the United Kingdom, Playfair was almost completely forgotten until
1861, when William Stanley Jevons enthusiastically adopted Playfair’s
methods in his own economic atlas (noted by Keynes, 1936). Jevons
(1886) wrote, “in statistics, the [graphical] method, never much used,
has fallen almost entirely into disuse. It ought, I consider, to be almost as
much used as maps are used in geography” (emphasis in original). Iron-
ically, Jevons never succeeded in publishing his economic atlas, styled in
the fashion of Playfair. Nonetheless, Jevons’s advocacy of the graphical
method found sympathetic ears in the British statistical establishment.
Most influential among those influenced by Jevons was Karl Pearson,
who not only embraced graphs, but included a lecture on charting in
his famous series of statistical lectures at University College, London.
Pearson acknowledged Playfair’s contributions in generous terms.
In the 20th century, the use of graphs increased markedly and text-
books soon appeared. Brinton (1914) may have been the first widely
sold primer on statistical graphs, but his text was quickly followed by
ahost of imitators. Today, it is normal to find graphs in newspapers,
magazines, periodicals, and professional journals to communicate quan-
titative phenomena; and other visual media, such as television, make
widespread use of charts for the same reason. Graphs are also used to
explore and analyze data: Statistical charting is an integral part of al-
most all computer software used in the sciences and commerce. Playfair
was well aware that charts were not merely a new way of presenting
data to others. He recognized that graphs could stimulate new ideas or
suggest models. After making a trial chart of some data, he said that
the first rough draft [gave] me a better comprehension of the sub-
ject, than all that I had learnt from occasional reading, for half my
lifetime” (Playfair, 1805, xv–xvi). Playfair charted data to discover as
well as to present; in that respect, he anticipated the exploratory uses of
graphs that were to become popular at the beginning of the 20th century
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
32 Introduction
(Spence & Garrison, 1993). William Playfair’s vision, which he was un-
able to communicate to others during his life, affects and benefits us all.
If he could see how his inventions have changed the ways in which we
analyze and present data, he would be enormously proud.
Ian Spence
Howard Wainer
BIBLIOGRAPHY: VARIATIONS OF
THE ATLAS AND BREVIARY
TheCommercial and Political Atlas
PRIVATELY CIRCULATED PRELIMINARY EDITION
Playfair, William (1785). The commercial and political atlas: Representing, by means of
stained copper-plate charts, the exports, imports, and general trade of England, at a single
view. London.
FIRST EDITION
Playfair, William (1786). The commercial and political atlas; representing, by means of
stained copper-plate charts, the exports, imports, and general trade of England, at a single
view.Towhich are added, Charts of the revenue and debts of Ireland, done in the same
manner by James Corry. London: Debrett; Robinson; and Sewell.
SECOND EDITION
Playfair, William (1787). The commercial, political, and parliamentary atlas, which repre-
sents at a single view, by means of copper plate charts, the most important public accounts
of revenues, expenditures, debts, and commerceofEngland. To which are added charts
of the revenues and debts of Ireland, done in the same manner, by James Corry, Esq.
London: Stockdale.
TRANSLATION OF THE SECOND EDITION
Playfair, William (1789). Ta bleaux d’arithm´etique lin´eaire, du commerce, des finances, et
de la dette nationale de l’Angleterre. Translated by Hendrik Jansen. Paris: Chez Barrois
l’aˆ
ın´
e.
LINEAL ARITHMETIC
Playfair, William (1798). Lineal arithmetic: Applied to shew the progress of the commerce
and revenue of England during the present century.London.
THIRD EDITION
Playfair, William (1801). The commercial and political atlas, representing, by means
of stained copper-plate charts, the progress of the commerce, revenues, expendi-
ture, and debts of England, during the whole of the eighteenth century.London:
Wallis.
TheStat istical Breviary
Playfair, William (1801). The statistical breviary: Shewing, on a principle entirely
new, the resources of every state and kingdom in Europe; illustrated with stained
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Introduction
copper-plate-charts ...to which is added, a similar exhibition of the ruling powers of
Hindoostan.London: Wallis.
Playfair, William (1802). ´
El´emens de statistique: o`u l’on d´emontre, d’apr`es un principe
enti`erement neuf, les ressources de chaque royaume, ´etat et r´epublique de l’Europe: suivis
d’un ´etat sommaire des principales puissances et colonies de l’Indostan. Orn´edecartes
colori´ees, repr´esentant, d’un coup-d’oeil, les forces physiques de toutes les nations Eu-
rop´eennes.Translated from the English by Denis Francois Donnant. Paris: Batilliot et
Genets.
REFERENCES
Author unknown (1823). Mr. William Playfair. Edinburgh Annual Register, 16, 332–4.
Belote, T.T. (1907). The Scioto speculation and the French settlement at Gallipolis.New
York:Burt Franklin
Beniger, J.R., & Robyn, D.L. (1978). Quantitative graphics in statistics: A brief history.
The American Statistician, 32, 1–11.
Biderman, A.D. (1981). The graph as victim of adverse discrimination and segregation:
Comment occasioned by the first issue of information design journal. Information
Design Journal, 1, 232–41.
Biderman, A.D. (1990). The Playfair enigma: The development of the schematic repre-
sentation of statistics. Information Design Journal, 6, 3–25.
Brinton, W.C. (1914). Graphic methods for presenting facts.NewYork:The Engineering
Magazine Company. Reprinted New York: Arno Press, 1980.
Broadie, A. (ed.) (2003). The Cambridge companion to the Scottish enlightenment.Cam-
bridge, UK: Cambridge University Press.
Buchan, J. (2003). Capital of the mind: How Edinburgh changed the world.Edinburgh:
Murray.
Cleveland, W.S. (1985). The elements of graphing data.Monterey,CA: Wadsworth.
Costigan-Eaves, P., & Macdonald-Ross, M. (1990). William Playfair (1759–1823). Sta-
tistical Science, 5, 318–26.
’Espinasse, M. (1962). Robert Hooke.Berkeley, CA: University of California Press.
Friendly, M. (2002). Visions and re-visions of Charles Joseph Minard. Journal of Educa-
tional and Behavioral Statistics, 27, 31–5.
Funkhouser, H.G. (1937). Historical development of the graphical representation of
statistical data. Osiris, 3, 269–404.
Funkhouser, H.G., & Walker, H.M. (1935). Playfair and his charts. Economic History (A
supplement to the Economic Journal), 3, 103–9.
Hankins, T.L. (1999). Blood, dirt, and nomograms: A particular history of graphs. Isis,
90, 50–80.
Herman, A. (2001). How the Scots invented the modern world: The tr ue story of how western
Europe’s poorest nation created our world and everything in it.NewYork:Crown.
Hooke, R. (1705). Of the true method of building a solid philosophy,or of a philosophical
algebra. In Richard Waller (ed.), Posthumous works, containing his Cutlerian lectures,
and other discourses, read at the meetings of the illustrious Royal Society.London.
Jefferson, T., & Peterson, M.D. (ed.) (1984). Writings: Autobiography / Notes on the State
of Virginia / Public and Private Papers / Addresses / Letters (Library of America).New
York: Library of America.
33
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Introduction
Jevons, H.A. (ed.) (1886). Letters and journal of W. Stanley Jevons.London: Macmillan.
Keynes, J.M. (1936). William Stanley Jevons 1835–1882: A centenary allocation on his
life and work as a statistician. Journal of the Royal Statistical Society, 99, 516–55.
Kosslyn, S.M. (1994). Elements of graph design.New York: Freeman.
Palsky, G. (1996). Deschiffres et des cartes, naissance et d´eveloppement de la cartographie
quantitative fran¸caise au XIXe siecle.Paris: Comit´
edes travaux historiques et scien-
tifiques.
Playfair, J. (1822). The works of John Playfair: With a memoir of the author.James G.
Playfair (ed.). Edinburgh: Constable.
Playfair,W. (1785). The increase of manufactures, commerce,and finance, w ith the extension
of civil liberty, proposed in regulations for the interest of money.London.
Playfair, W. (1805). An inquiry into the permanent causes of the decline and fall of powerful
and wealthy nations.London: Greenland & Norris.
Playfair, W. (1809–11). British family antiquity illustrative of the origin and progress of
the rank, honours, and personal merit, of the nobility of the United Kingdom.London:
Reynolds & W. Playfair.
Playfair, W. (1821). Aletter on our agricultural distresses, their causes and remedies: Accom-
panied with tables and copper-plate charts, shewing and comparing the prices of wheat,
bread, and labour, from 1565 to 1821.London: Sams.
Playfair, W. (1822). Aletter on our agricultural distresses, their causes and remedies: Accom-
panied with tables and copper-plate charts, shewing and comparing the prices of wheat,
bread, and labour, from 1565 to 1821,3rd ed., with an additional chart. London: Sams.
Playfair,W. (1822–3). Unpublished ms, held byJohn Lawrence Playfair, Toronto, Canada.
Transcribed and annotated by Ian Spence.
Priestley, J. (1765). A Chart of Biography.London: William Eyres.
Reid, T. (1764). An inquiry into the human mind, on the principles of common sense.
Edinburgh: Kincaid & Bell.
Schofield, R.E. (1963). The Lunar Society of Birmingham: A social history of provincial
science and industry in eighteenth-century England.Oxford: Clarendon Press.
Smith, A. (1805). An inquiry into the nature and causes of the wealth of nations,11th ed.,
with notes, supplementary chapters, and a life of Dr. Smith, by William Playfair.
London: Cadel and Davies.
Spence, I. (2000). The invention and use of statistical charts. Journal de la Soci´et´eFran¸caise
de Statistique, 141, 77–81.
Spence, I. (2004). William Playfair. Oxford dictionary of national biog raphy, Vol. 44, 562–3.
Oxford: Oxford University Press.
Spence, I. (2005). No humble pie: The origins and usage of a statistical chart. Journal of
Educational and Behavioral Statistics,30 (3).
Spence, I., & Garrison, R.F. (1993). A remarkable scatterplot. The American Statistician,
47, 12–19.
Spence, I., & Lewandowsky, S. (1990). Graphical perception, Ch. 1., pp. 13–57, in J. Fox
&J.S. Long (eds.), Modern methods of data analysis.Beverly Hills, CA: Sage.
Spence, I., & Wainer, H. (1997). William Playfair: A daring worthless fellow. Chance, 10,
31–4.
Spence, I., & Wainer, H. (2001). William Playfair (1759–1823): Inventor and ardent
advocate of statistical graphics. In C.C. Heyde, & E. Seneta (eds.), Statisticians of the
Centuries.New York: Springer-Verlag, 105–10.
34
P1: FpQ
0521855543int Playfair 0 521 85554 3 September 26, 2005 20:27
Introduction 35
Spence, I., & Wainer, H. (2004). William Playfair. In K. Kempf-Leonard (ed.), Encyclo-
pedia of Social Measurement.San Diego: Academic Press, 71–9.
Tilling, L. (1975). Early experimental graphs. British Journal for the History of Science, 8,
193–213.
Tufte, E.R. (1983). The visual display of quantitative information.Cheshire, CT: Graphics
Press.
Uglow, J. (2002). The lunar men: Five friends whose curiosity changed the world.London:
Faber&Faber.
Valois, J.-P. (2000). L’approche graphique en analyse des donn´
ees (avec discussions).
Journal de la Soci´et´eFrancaise de Statistique, 141, 5–40.
Wainer, H. (2000). Visual revelations: Graphical tales of fate and deception from Napoleon
Bonaparte to Ross Perot,2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates.
Wainer, H. (2005). Graphic discovery: A trout in the milk and other visual adventures.
Princeton, NJ: Princeton University Press.
Wainer, H., & Velleman, P.F. (2001). Statistical graphics: Mapping the pathwaysof science.
Annual Review of Psychology, 52, 305–35.
Wainer, H., & Spence, I. (2005). Graphical presentation of longitudinal data. In B. Everitt
&D.C. Howell (eds.), Encyclopedia of Behavioral Statistics.NewYork:Wiley.
... Evidenziare il dato centrale Come attuato già da Playfair nel 1786, si possono incorporare elemen­ ti informativi e/o metaforici nella visualizzazione, che possano facilitarne l'interpretazione e la comprensione ( Playfair et al. 2005). Quello che infatti differenzia un progetto di visual journalism da una infografica è l'inserimento di questa all'interno di una narrativa più ampia. ...
... Gli errori grafici: l'effetto tridimensionale Rimuovere ombre o bordi è vivamente consigliato, non aggiunge nes­ sun beneficio interpretativo ( Figura 5). Un altro dettaglio importante che facilita la lettura di una visualiz­ zazione è quello dell'incorporamento (o sovrapposizione) della legenda, tecnica già adottata da Playfair ( Playfair et al. 2005) qualche secolo fa. Apparentemente un dettaglio, facilita enormemente l'esperienza infor­ mativa, evitando all'occhio del lettore continui salti all'interno del lavo­ ro in cerca di informazioni ( Figura 6). ...
... Graphical representation dates back more than 200 years (Unwin, 2008) For example, (Wainer & Spence, 2005) compiled the work of (Playfair, 1801) who represented data in graphical form. The discipline of statistics uses either real or hypothetical data which can be interpreted graphically (Cleveland, 1985;Tufte, 2001). ...
Article
Full-text available
Recently, statistical reasoning has been of vital importance not only in quantitative analysis but also in the interpretation of graphs at all educational levels. There are students that can make calculations almost immediately but are not able to interpret or present their ideas graphically. In this way, the present study seeks to conduct a diagnostic of the problems that economic-administrative students have when reading and interpreting graphs in their statistics courses. For this, a Spanish version of the test Comprehensive Assessment of Outcomes in Statistics (CAOS) was administered. This instrument allows for the determination of reasoning applied to different types of statistical graphs and in some cases to determine what type of calculation is required to do it. The instrument was applied to 138 undergraduate students from the economic-administrative area of the University of Guadalajara during January-June 2018. The results show that a large percentage of students confuse a normal distribution with a uniform one and that they are unable to distinguish that a bias can be determined from the measures of central tendency and dispersion, as well as other statistical reasoning difficulties. This may be as a result of a deficiency that exists in statistical teaching, an insufficient mathematical preparation on the part of the students, among other factors. https://www.ride.org.mx/index.php/RIDE/article/view/786
... From [79] plot known from the tenth or eleventh century showing the inclination of the planets as a function of time [90], but this way of presenting information appears to have been a oneoff and did not influence the way numerical information was presented in the following centuries). Several people have suggested reasons why the graphical way of presenting numerical information was not widely used until the middle of the nineteenth century [91][92][93]. The problem of presenting data in tabular form was well-expressed by Playfair in 1801 [94] (p. ...
Article
Full-text available
Up until the Industrial Revolution, the dynamic mechanical properties of materials were only of importance in warfare, particularly after the powder-driven gun was invented. With the invention of the steam engine, the explosion of steam boilers (which is similar to the explosion of cannon) became a concern. When railways began to be built, the lack of knowledge of the dynamic properties of the iron alloys used in rails and railway bridges was understood to be a problem, but no way of measuring them was devised until the end of the nineteenth century. Ingenious mechanical (and later electromechanical) methods of recording signals onto rotating drums or moving smoked glass plates began to be developed from the middle of the nineteenth century onwards. Optical/photographic methods of recording information from dynamic experiments date from the 1890s. The rod-on-anvil technique (later named after Taylor) was developed in France at the beginning of the twentieth century but not mathematically analysed until the 1940s. The Hopkinson pressure bar was invented just before the start of the First World War and found to be useful in improving British artillery shells. It was then forgotten about until the Second World War when a two-bar version was developed for measuring the dynamic properties of soft materials such as explosives and polyethylene. As the story of high rate mechanical testing from about 1950 onwards is quite well known to the high rate testing community, this date is taken as the end point of this article.
... The cornerstone of Parallel Bubbles is a "bubble" (a circle) of variable radius that represents a categorical level. It was inspired by the "bubble plot" invented more than two centuries ago by Playfair (1801). A similar feature has already been available in Parabox by Advisor Solutions and is presented in Few (2006). ...
Conference Paper
Full-text available
Parallel Coordinates are a widely used visualization method for multivariate data analysis tasks. In this paper we discuss the techniques that aim to enhance the representation of categorical data in Parallel Coordinates. We propose Parallel Bubbles, a method that improves the graphical perception of categorical dimensions in Parallel Coordinates by adding a visual encoding of frequency. Our main contribution consists in a user study that compares the performance of three variants of Parallel Coordinates, with similarity and frequency tasks. We base our design choices on the literature review, and on the research guidelines provided by Johansson et al (2016). Parallel Bubbles are a good trade-off between Parallel Coordinates and Parallel Sets in terms of performance for both types of tasks. Adding a visual encoding of frequency leads to a significant difference in performance for a frequency-based task consisting in assessing the most represented category. This study is the first of a series that will aim at testing the three visualization methods in tasks centered on the continuous axis, and where we assume that the performance of Parallel Sets will be worse.
... The genius of Playfair was to overcome that difficulty through the invention of graphs and statistical charts. In the explanation of his invention he warned the reader [1]: "In studying each of these diagrams carefully, a sufficiently distinct impression will be made, to remain abandoned for a considerable time, and the idea that remains will be simple and complete, enclosing together time duration and dimension. People can only focus on the general vision; the information will be read without effort or problems studying the details of which it is constituted". ...
Conference Paper
Full-text available
Graphic representation is one of the main systems of signs designed by man with the aim of preserving, understanding, and communicating information deemed essential. As a language for eyes, graphic representation benefits from the properties of visual perception. The effectiveness of graphic language has been recognized and studied by many theorists and communication professionals. The contribution represents in the form of notes some initial explorations in the field of infographic and graphic visualization as knowledge processing tools for analyzing and present data as information. There are some remarkable references, definitions and examples, of researchers and information professionals, from which there are extensive considerations that attempt to bring the principles of visual communication to the science of representation and to the drawing discipline for engineers and architects. Keywords: infographics; graphic visualization; visible variables; interaction project.
... The power of using graphs to facilitate decision-making has long been valued by those charged with making critical decisions. King Louis XVI described graphs as speaking all languages (Playfair, 2005). To this day, visual displays are thought to be an efficient and beneficial method of presenting data for decision-making. ...
Article
The quality of a data display can have an impact on the interpretation of those data. A survey of the literature indicates that data displays can vary in quality of accuracy, clarity, and efficacy. In this study we develop and apply an evaluative rubric to graphs in a sample of six education journals: three research and three practitioner. Results indicate that graph quality is typically high in educational journals, however, in practitioner oriented journals issues around graph clarity and efficacy should be addressed. Common error patterns are pinpointed, and four recommendations are made to authors and editors: focus on meaningful labels, increase amount of data displayed, portray multiple relationships, and elaborate with supporting text.
Conference Paper
Full-text available
Scatter plots are a powerful and well-established technique for visualizing the relationships between two variables as a collection of discrete points. However, especially when dealing with large and dense data, scatter plots often exhibit problems such as overplotting, making the data interpretation arduous. Density plots are able to overcome these limitations in highly populated regions, but fail to provide accurate information of individual data points. This is particularly problematic in sparse regions where the density estimate may not provide a good representation of the underlying data. In this paper, we present sunspot plots, a visualization technique that communicates dense data as a continuous data distribution, while preserving the discrete nature of data samples in sparsely populated areas. We furthermore demonstrate the advantages of our approach on typical failure cases of scatter plots within synthetic and real-world data sets and validate its effectiveness in a user study.
Article
Information visualization is a rapidly evolving field with a growing volume of scientific literature and texts continually published. To keep abreast of the latest developments in the domain, survey papers and state‐of‐the‐art reviews provide valuable tools for managing the large quantity of scientific literature. Recently, a survey of survey papers was published to keep track of the quantity of refereed survey papers in information visualization conferences and journals. However, no such resources exist to inform readers of the large volume of books being published on the subject, leaving the possibility of valuable knowledge being overlooked. We present the first literature survey of information visualization books that addresses this challenge by surveying the large volume of books on the topic of information visualization and visual analytics. This unique survey addresses some special challenges associated with collections of books (as opposed to research papers) including searching, browsing and cost. This paper features a novel two‐level classification based on both books and chapter topics examined in each book, enabling the reader to quickly identify to what depth a topic of interest is covered within a particular book. Readers can use this survey to identify the most relevant book for their needs amongst a quickly expanding collection. In indexing the landscape of information visualization books, this survey provides a valuable resource to both experienced researchers and newcomers in the data visualization discipline.
Thesis
Visualizations in the most general sense of external, physical representations of information are older than the invention of writing. Generally, external representations promote external cognition and visual thinking, and humans developed a rich set of skills for crafting and exploring them. Computers immensely increased the amount of data we can collect and process as well as diversified the ways we can represent it visually. Computer-supported visualization systems, studied in the field of information visualization (infovis), have become powerful and complex, and sophisticated interaction techniques are now necessary to control them. With the widening of technological possibilities beyond classic desktop settings, new opportunities have emerged. Not only display surfaces of arbitrary shapes and sizes can be used to show richer visualizations, but also new input technologies can be used to manipulate them. For example, tangible user interfaces are an emerging input technology that capitalizes on humans' abilities to manipulate physical objects. However, these technologies have been barely studied in the field of information visualization. A first problem is a poorly defined terminology. In this dissertation, I define and explore the conceptual space of embodiment for information visualization. For visualizations, embodiment refers to the level of congruence between the visual elements of the visualization and their physical shape. This concept subsumes previously introduced concepts such as tangibility and physicality. For example, tangible computing aims to represent virtual objects through a physical form but the form is not necessarily congruent with the virtual object. A second problem is the scarcity of convincing applications of tangible user interfaces for infovis purposes. In information visualization, standard computer displays and input devices are still widespread and considered as most effective. Both of these provide however opportunities for embodiment: input devices can be specialized and adapted so that their physical shape reflects their functionality within the system; computer displays can be substituted by transformable shape changing displays or, eventually, by programmable matter which can take any physical shape imaginable. Research on such shape-changing interfaces has so far been technology-driven while the utility of such interfaces for information visualization remained unexploited. In this thesis, I suggest embodiment as a design principle for infovis purposes, I demonstrate and validate the efficiency and usability of both embodied visualization controls and embodied visualization displays through three controlled user experiments. I then present a conceptual interaction model and visual notation system that facilitates the description, comparison and criticism of various types of visualization systems and illustrate it through case studies of currently existing point solutions. Finally, to aid the creation of physical visualizations, I present a software tool that supports users in building their own visualizations. The tool is suitable for users new to both visualization and digital fabrication, and can help to increase users' awareness of and interest in data in their everyday live. In summary, this thesis contributes to the understanding of the value of emerging physical representations for information visualization.
Article
Full-text available
Ian Spence, Colin R. Fenn and Scott Klein go in search of the final resting place of William Playfair, who devised many of the statistical graphics in use today
Article
Full-text available
Charles Joseph Minard is most widely known for a single work—his poignant flow-map depiction of the fate of Napoleon’s Grand Army in the disastrous Russian campaign of 1812. In fact, Minard was a true pioneer in thematic cartography and in statistical graphics; he developed many novel graphic forms to depict data, always with the goal to let the data “speak to the eyes.” This article reviews Minard’s contributions to statistical graphics, the time course of his work, and some background behind the famous March on Moscow graphic. This article also looks at some modern revisions of this graph from an information visualization perspective and examines some lessons this graphic provides as a test case for the power and expressiveness of computer systems or languages for graphic information display and visualization.
Article
Full-text available
William Playfair's pie chart is more than 200 years old and yet its intellectual ori-gins remain obscure. The inspiration likely derived from the logic diagrams of Llull, Bruno, Leibniz, and Euler, which were familiar to William because of the instruction of his mathematician brother John. The pie chart is broadly popular but—despite its common appeal—most experts have not been seduced, and the academy has advised avoidance; nonetheless, the masses have chosen to ignore this advice. This commentary discusses the origins of the pie chart and the appro-priate uses of the form.
Article
L. J. Henderson, a Harvard physiologist and the first president of the History of Science Society, attempted to analyze mammalian blood solely as a physical-chemical substance. He found that the only way he could describe a chemical system as complicated as blood was by a diagram called a "nomogram." This lecture tells the history of Henderson's nomogram and of nomograms in general. It describes the origins of graphs in the eighteenth century, their development in nineteenth-century engineering practice, and their importance in the twentieth century for describing physical and chemical systems.
Article
Constraints of typography have led to the physical segregation of diagrammatic treatments of information from the linear alphanumerics that have dominated scientific communication. Information Design Journal reflects these constraints. The segregated and subordinated status of graphics is metaphorically compared here to the cultural roles of women and minorities. Aspects of graphics treated include the retardation of their intellectual development and influence; their relegation to relatively tangential, simplistic and ornamental functions; and prevalent suspicions regarding their deceptiveness. However, technology now makes it possible to reintegrate graphics into the mainstream of intellectual communication.
Article
The invention of statistical graphics is generally, if inaccurately, attributed to William Playfair His initial innovation, along with his subsequent invention of most of the major repertoire of statistical graphics, is in many ways an enigma of the history of science: (1) Given their apparent obviousness, why had these graphic forms not been previously used for plotting statistics? {2} Why was the Cartesian coordinate system, during a century ami a half from its invention, not regularly applied to the kinds of data which Playfair plotted? (3) Why were the symbolic schematics used by Playfair apparently understood by contemporaries without need for prior learning of his 'conventions'? (4) Why did serious scholarly attention to Playfair'$ innovations occur earlier on the continent than in England? (5) Why subsequently have there been waves of popularity and of neglect of Playfair's forms? (S) Why were statistical graphics invented by a political pamphleteer and business adventurer rather than a scholar or scientist? (7) Why did statistical graphics develop first for social data applications rather than for natural or physical science purposes? Addressing these questions may shed light on developments in schematic representation of statistics from the beginnings of cultural numeracy to the present day The primary explanations of the enigma are: (1) the similarities and differences between the purely empirical data graph and diagrammatic representations of pure or applied mathematical functions; (2) the association of utility of pure data graphs with a statistical orientation toward phenomena, Playfaiťs innovations were facilitated by bis association with science during a time when science was particularly hospitable to highly pragmatic endeavors. His innovations were also facilitated by bis marginality with regard to the science of bis contemporaries.
Article
The graphical presentation of experimental data in the physical sciences has several advantages which today are too familiar to require very detailed enumeration. Its greatest strength lies in the clarity and succinctness with which it displays the information contained in tabulated results: for the experimenter a graph provides a rough and immediate check on the accuracy and suitability of the methods he is using, and for the reader of a scientific report it may convey in a few seconds information that could only be gleaned from a table of measurements by hours of close study. There are occasions where only the analysis of experimental graphs will provide the information we require, but usually the actual analysis of results is carried out nowadays by computational methods. The use of graphs is therefore not so much a necessary part of scientific procedure as an extremely useful one, and one that is often taken very much for granted.
A distinguished team of contributors examines the writings of David Hume, Adam Smith, Thomas Reid, Adam Ferguson, Colin Maclaurin and other Scottish thinkers, in philosophy, natural theology, economics, anthropology, natural science and law. The contributors also relate the Scottish Enlightenment to its historical context and assess its impact and legacy in Europe, America and beyond. The volume is of interest to a wide range of readers in philosophy, theology, literature and the history of ideas.