History on the Line:
Time as Dimension
Stephen Boyd Davis

1 Stephen Ferguson, “The 1753
Introduction
Carte Chronographique of Jacques
In 1753, Jacques Barbeu-Dubourg wrote a pamphlet accompanying
Barbeu-Dubourg,” Princeton University
a timeline of history, his Chronography or Depiction of Time.1 In it he
Library Chronicle (Winter 1991).
compared the visualization of past time unfavorably with the visu-
http://www.princeton.edu/~ferguson/
PULC_1991_duBourg.pdf (accessed
alization of distant countries. Geography he found easy, attractive,
December 17, 2011).
and highly memorable while chronology was “laborious, unforgiv-
2 Jacques Barbeu-Dubourg, Chronographie,
ing, offering nothing to the mind but repel ent dates.”2 A solution
ou Description des Tems [Chronography,
seemed to present itself:
or Depiction of Time], trans. Stephen

Geography has as its object the extent of the earth;
Boyd Davis (Paris, 1753 [2009]), 5.
http://goo.gl/vNhN (accessed December

Chronology has as its object the succession of time. May
17, 2011).

not duration be imitated and represented as effectively
3 Ibid.

to the senses as distinctly as space, and may not intervals
4 Jacques Barbeu-Dubourg, Chronographie

of time be as easily counted in degrees?3
Universelle & Détails qui en Dépendent
pour la Chronologie & les Généalogies
(Paris: Barbeu-Dubourg, Lamote, Fleury,
In this way Barbeu-Dubourg offered his rationale for one of the
1753). Princeton University Library Rare
earliest timelines: a representation mapping time arithmetical y to
Books: D11 .B37 1753.
graphical space rather than presenting events in a list or table
5 Joseph Priestley, A Chart of Biography
(see Figure 1).4 His publication was followed in 1765 by Joseph
(London, 1765). British Library: 611.l.19.
Priestley’s Chart of Biography (see Figure 2).5 One of Priestley’s inno-
6 The two inventors shared a connection:
Benjamin Franklin was a mutual
vations was the use of actual graphic lines to represent the life
friend and sent Barbeu-Dubourg a
duration of individuals.6
copy of Priestley’s timeline, to which

In some ways, little has changed since Barbeu-Dubourg
Barbeu-Dubourg responded with the
contrasted the state of the art in geographic and temporal visual-
gracious remark, “I have received with
ization. Admittedly, many charts—some of them sophisticated—
gratitude and viewed with pleasure the
biographical chart of Mr. Priestley which
have mapped quantities against time: Quantitative visualization
is in truth made on almost the same prin-
took off shortly after Barbeu-Dubourg and Priestley’s innovations
ciples as my own, without plagiarism on
and was partly based on their work. But the visualization of his-
either side, as I in no way claim primacy
toric lives, events, and artifacts and the relations between them—
on account of the date.” He may have
known from Franklin that Priestley had
of categorical rather than quantitative information about the
been making charts for some time to
past—is rarely given serious consideration as either a design or
assist in his teaching at the Warrington
historiographic problem. As a result of long and intense research
Dissenting Academy before he began to
and development, geographic maps do their job wel , and a wealth
publish them. (Jacques Barbeu-Dubourg
of experience is applied to their design, supported in these
to Benjamin Franklin, May 8, 1768, in
The Papers of Benjamin Franklin,
days by significant innovations arising from digital capabilities.
http://franklinpapers.org/franklin/yale?vo
Controversy rages over the respective merits of the Mercator,
l=15&page=112a&ssn=001-78-0029
Gall-Peters, and other projections, with a clear understanding
(accessed December 21, 2011).
© 2012 Massachusetts Institute of Technology
4
DesignIssues: Volume 28, Number 4 Autumn 2012

---

Figure 1
Detail of Barbeu-Dubourg’s 1753 Carte
Chronographique. Time runs from left to right.
The other axis is used for states and king-
doms, with two bands near the bottom edge
for notable events and individuals. This detail
shows the lowest band for the period from
about 1693AD to 1728AD. Rare Book Division,
Department of Rare Books and Special
Collections, Princeton University Library.
Reproduced with permission.
Figure 2
Priestley’s 1765 Chart of Biography (detail).
Time extends from left to right. This detail
shows the period 1300AD to 1800AD. The
other axis is divided into horizontal bands,
here Mathematicians and Physicians.
Reproduced with the permission of Chetham’s
Library, Manchester.
7 See Denis Wood and John Krygier,
that each presents a different world view and that these differ-
“Critical Cartography,” in International
ences matter.7 The mapping of time, by contrast, has made only
Encyclopedia of Human Geography, eds.
modest intel ectual progress since it was invented 250 years ago:
Rob Kitchin and Nigel Thrift (Amsterdam:
The choices to be made by the designer are assumed to be straight-
Elsevier, 2009), 340-44.
forward and incite no debate. Twyman has pioneered timeline
8 See Michael Twyman, “Articulating
Graphic Language: A Historical
scholarship from a typographic and graphical point of view, while
Perspective,” in Towards a New
the recent survey by Rosenberg and Grafton is scholarly, compre-
Understanding of Literacy, eds.
hensive, and highly visual.8 Rosenberg’s extensive article on the
Merald E. Wrolstad and Dennis Fisher
timelines of Joseph Priestley is also a major contribution to the
(New York: Praeger, 1986), 188–251;
subject.9 However, this fine scholarship has been applied to recov-
Michael Twyman, “Textbook Design:
Chronological Tables and the Use of
ering the lost history of chronographics rather than to extracting
Typographic Cueing,” Paradigm 4
key principles that might help us design new ones. I hope to make
(December 1990); and Daniel Rosenberg
the latter contribution in this article. Doing so involves asking
and Anthony Grafton, Cartographies
what we are mapping and how. I draw on psychological and
of Time: A History of the Timeline
anthropological sources and on my own historical investigations
(New York: Princeton Architectural
into early timelines and their authors. Crudely, one could say that
Press, 2010).
9 Daniel Rosenberg, “Joseph Priestley and
the eighteenth-century pioneers had subtle ideas but not the tech-
the Graphic Invention of Modern Time,”
nology to implement them, while now we have technologies that
Studies in Eighteenth Century Culture 36
they sorely needed but less critical thought. Apart from imaging
no. 1 (2007): 59.
DesignIssues: Volume 28, Number 4 Autumn 2012
5

---

the past, timelines are an important component of many digital
tools for planning projects—for creating music, animations, mov-
ies, and multimedia—and for representing processes at a range of
scales from nanoseconds to the geological and cosmological. If we
can think intel igently about mapping time to graphical space in
the domain of history, this capacity will benefit other fields as wel .
Time as Dimension
This article focuses on one issue—the orientation and direction
of the time axis—to the exclusion of many others; for example, it
touches only briefly on the closely related questions of scale and
calibration. In addition, even the earliest chronologies showed a
strong awareness of the usefulness of being able to look at simulta-
neous events, “across” time, as well as to displaying their date
order, but the design and use of these non-time axes is a topic
for another day. What fol ows is also confined to mapping time to
a straight line. This model is of course just one possibility, often
contrasted with cyclic or circular models allegedly held by, for
example, Greek, Asian, or any non-Judaeo-Christian cultures. Such
binary contrasts have been rejected as simplistic by Feeney, Goody,
and Gould.10 In fact, in any culture we seem to operate comfortably
with switchable concepts of time. In the words of Möller and
10 See Denis Feeney, Caesar’s Calendar:
Luraghi, “most people perceive time in different ways according to
Ancient Time and the Beginnings of
their context or situation, with the result that any one culture is
History (Berkeley: University of California
characterised by a range of different perceptions of time.”11
Press, 2007), 3; Jack Goody, The Theft of
History (Cambridge UK: Cambridge
University Press, 2006), 18; and Stephen
The Uses of Timelines
Jay Gould, Time’s Arrow, Time’s Cycle:
We cannot begin to look at the form of the timeline without con-
Myth and Metaphor in the Discovery of
sidering its purpose. What is the objective when historical time is
Geological Time (Cambridge, MA:
mapped on a surface—whether page, exhibition board, or digital
Harvard University Press, 1987).
display? What are such maps of time meant to achieve? Potential y,
11 Astrid Möller and Nino Luraghi, “Time in
the Writing of History: Perceptions and
they share the benefits of any wel -designed data visualization,
Structures,” Storia della Storiografia 28
enabling users to spot patterns, trends, clusters, gaps, and outli-
(1995), 7.
ers—in short, to make sense of data.12 Visualizations are not only
12 For a useful synthesis of the literature,
for the benefit of other people; authors themselves might also ben-
see Stuart K. Card, Jock D. Mackinlay,
efit. These representations serve the social reproduction of knowl-
and Ben Shneiderman, eds., Readings
in Information Visualization: Using Vision
edge and they constitute tools for thinking.13 When applied to
to Think (San Francisco: Morgan
historic data, specific insights include spotting connections
Kaufmann, 1999), 16.
through and across time, identifying clusters and lacunae, and
13 Matthias Schemmel, “Medieval
understanding the context of individual events, actions, and arti-
Representations of Change and Their
facts. Such benefits have been claimed from the earliest days of
Early Modern Application,” Preprint 402.
use. Priestley argued that users would be able to see how individu-
TOPOI – Towards a Historical
Epistemology of Space (Max Planck
als “stand related in point of time to one another,” get “a clearer
Institute for the History of Science, 2010),
idea of the time in which they lived,” understand “the state of their
www.mpiwg-berlin.mpg.de/Preprints/
co[n]temporaries,” assess the “relative length of their lives,” and
P402.pdf (accessed December 17, 2011).
perceive “the intervals of time which elapsed between them and
14 Joseph Priestley, A Description of a Chart
their predecessors and successors.”14 A subtle and insightful man,
of Biography (Warrington: Eyre, 1764),
British Library: 611.d.30, 3.
6
DesignIssues: Volume 28, Number 4 Autumn 2012

---

Figure 3
A timeline as public exhibit in the Cartoon
Museum, London, UK, 2007. In this example,
the items are packed, with no attempt to plot
events arithmetically. Sequence is clearly
represented, but interval and scale are not.
Homer Creative Limited, Birmingham. Used
with permission. Photo: Stephen Boyd Davis.
Priestley realized that gaps are meaningful: “The thin and void
places in the chart are, in fact, not less instructive than the most
crowded.” The combination of detail and context was important to
him: “It is a peculiar kind of pleasure we receive, from such a view
as this chart exhibits, of a great man, such as Sir Isaac Newton,
seated, as it were, in the circle of his friends and il ustrious co[n]-
temporaries. We see at once with whom he was capable of holding
conversation….”15 All this would be “particularly useful to stu-
dents in Chronology, History, and Biography.”16
The Emergence of Modern Chronographics
In the mid-eighteenth century, a shift from typographic, tabular
layouts of events to truly graphical time-maps il ustrated a chang-
ing model of time exemplified by the ideas of Descartes and
15 Ibid., 24.
Newton.17 The event-based model meant that earlier chronologies
16 Ibid., 4.
had almost always packed each tabular entry close upon the previ-
17 Stephen Boyd Davis, Emma Bevan, and
ous one. Blair complained that they have “all of them made one
Aleksei Kudikov, “Just in Time: Defining
great and fundamental Mistake […] contracting History into as
Historical Chronographics,” in EVA
London 2010: Electronic Visualisation and
little Room as they could….”18 Admittedly, Helvicus had created
the Arts:Proceedings of a Conference
tables using equal space—in his case, equal pages—for equal time,
Held in London (London: British
but another 140 years elapsed before historic time was mapped
Computer Society, 2010), 355-362. www.
arithmetically to space on a continuous substrate, rather than
bcs.org/content/conWebDoc/36111
table-wise in pages.19 Just like those pre-eighteenth century tables,
(accessed December 17, 2011).
the majority of current “timelines” are real y lists in which each
18 John Blair, The Chronology and History of
the World, from the Creation to the Year
item is allotted just enough space for its name or description—
of Christ, 1753 (London, 1754). British
especially on the Web, where the limitations of early versions
Library 1852.c.9.
of HTML tended to encourage the production of packed lists and
19 Christophorus Helvicus [Christoph
tables rather than anything graphically sophisticated. Event
Helvig], Theatrum Historicum [Historical
lists can be given a graphical treatment, as shown in Figure 3, but
and Chronological Theatre], (Giessæ
Hessorum, 1609). British Library 747.c.22.
DesignIssues: Volume 28, Number 4 Autumn 2012
7

---

time here is just an ordering principle rather than a dimension.
Such treatments convey information about sequence but not about
interval or scale.
Time as Dimension: Orientation and Direction
Whether we can think about time without using metaphors from
other domains is questionable.20 Gentner emphasizes the sheer util-
ity of mapping time to space: Location captures elements and their
relations; dimension shows duration; it is an eminently usable ana-
logue of an abstract concept.21 However, if time is mapped to a line,
on which axis of the graphic surface should it lie?22 And in which
Figure 4
direction should later times lie in relation to earlier: What is the
Axes relative to the perceiving subject,
direction of travel? Can we identify an answer to these questions
labeled according to their relation to the
that is the most natural, the best, or simply right?
body and with the conventional Cartesian

Often in tabular layouts of history, two directions are used
x, y, and z.
together to represent successive times. In the West, this orientation
is usual y left-to-right and top-to-bottom, fol owing standard read-
20 Lynn Avery Hunt, Measuring Time,
ing practice—characterized by Twyman as “linear interrupted.”23
Making History (Budapest: Central
The page is divided into vertical y sequenced rows representing
European University Press, 2008),
events or years, while successive pages, turned horizontal y, repre-
3 passim.
21 Dedre Gentner, “Spatial Metaphors
sent larger increments of time, as in a diary. When time began to
in Temporal Reasoning,” in Spatial
be ful y mapped to space, there were thus already tabular prece-
Schemas and Abstract Thought, ed.
dents for considering both left-to-right and top-to-bottom as the
Merideth Gattis (Cambridge, MA:
“right” way for time’s flow. But beyond the graphical domain, lan-
MIT Press, 2003), 203-22 at 221.
guage and gesture both give clues as to how time may be con-
22 Tversky cites a number of arguments
for preferring the cardinal axes, to which
ceived or experienced as directional. Interestingly, the evidence
the present discussion is confined. See
from language and the evidence from gesture are in conflict.
Barbara Tversky, “Spatial Schemas in
Depictions” in Spatial Schemas and
Direction in Verbal and Gestural Metaphors for Time
Abstract Thought, ed. Merideth Gattis
The significance of verbal metaphors was noted by Priestley, who
(Cambridge, MA: MIT Press, 2003),
argued that line-length is an appropriate representation because
79-112 at 99.
23 Michael Twyman, “A Schema for the
both lines and time can be described using the words “short” and
Study of Graphic Language,” in Media,
“long.”24 He did not discuss the appropriateness of any particular
Knowledge and Power, ed. Oliver Boyd-
direction. Traugott was one of the earliest researchers to ask why
Barrett and Peter Braham (London: Croom
verbal metaphor prefers some directions to others.25 Lakoff and
Helm, 1987), 201-25.
Johnson cite examples in English where the future lies in front of
24 Priestley,
A Description of a Chart of
Biography (1764), 7.
the observer (e.g., “the weeks ahead of us”), where time is moving
25 Elizabeth C. Traugott, “Spatial
toward the observer, and where the observer is moving through
Expressions of Tense and Temporal
time.26 All these are sagittal metaphors: Time is aligned front to
Sequencing: A Contribution to the Study
back relative to the observer’s body (see Figure 4). Alverson found
of Semantic Fields,” Semiotica 15 no. 3
that metaphors of the observer moving through static time and of
(1975): 207-30.
time flowing around the observer occur in many languages.27 Not
26 George Lakoff and Mark Johnson,
Metaphors We Live By (Chicago:
only is the sagittal axis preferred in verbal metaphor, but so is one
University of Chicago Press, 1996), 41-45.
direction along that axis. Núñez and Sweetser conclude that al
27 Hoyt Alverson, Semantics and
documented languages, with the apparent exception of Aymara in
Experience: Universal Metaphors of Time
the Andes, map future events onto spatial locations in front of
in English, Mandarin, Hindi, and Sesotho
(Baltimore: Johns Hopkins University
Press, 1994).
8
DesignIssues: Volume 28, Number 4 Autumn 2012

---

speakers, with past events behind them;28 however, evidence has
recently emerged of another rare model that orients time-ordered
objects not relative to the observer but relative to the world.29

Although verbal metaphors for time are dominated by the
sagittal, some evidence of vertical language concepts also emerges.
Mandarin describes earlier events as “shàng” or “up,” while later
events are “xià” or “down.” French uses the terms remonter (travel
(back) up) for looking back to early events, and descendre (descend)
for passing down the generations. English, of course, uses the
28 Rafael E. Núñez and Eve Sweetser,
French-derived word descendant. Boroditsky, Fuhrman, and
“With the Future Behind Them:
Convergent Evidence from Aymara
McCormick suggest that Mandarin has more vertical examples
Language and Gesture in the Cross-
than English because the language is traditional y written verti-
Linguistic Comparison of Spatial
cal y, top to bottom.30 However, even in Mandarin, horizontal met-
Construals of Time,” Cognitive Science
aphorical use outnumbers vertical.
30 no. 3 (2006): 401-50.

Work on metaphoric gesture adds weight to the evidence
29 Lera Boroditsky and Alice Gaby,
for spatial and directional models of time. However, it also adds
“Remembrances of Times East: Absolute
Spatial Representations of Time in an
a complication. We have seen that language metaphors for time
Australian Aboriginal Community,”
are sagittal or vertical. Verbal left-right expressions for time are
Psychological Science 21 no. 11 (2010):
“virtually non-existent,”31 and we do not say that someone did X
1635-39.
“to the left of” Y to mean that someone did X before Y.32 How-
30 Lera Boroditsky, Orly Fuhrman and Kelly
ever, gestural metaphors are often lateral. A detailed study of one
McCormick, “Do English and Mandarin
Speakers Think About Time Differently?”
French speaker by Calbris reveals extensive use of the left and
Cognition (2010). http://psych.stanford.
right hands to denote the past and future, respectively, which she
edu/~lera/papers/mandarin-time-2010.
believes is inspired by reading direction.33 Cienki also reports a
pdf (accessed December 19, 2011).
gestural pattern with prior time to the left and later time to the
31 Traugott, “Spatial Expressions of Tense
right and agrees on the influence of writing.34 Cooperrider and
and Temporal Sequencing,” 214.
Núñez discerned five distinct temporal gestures, in all of which
32 Alan J. Cienki and Cornelia Müller,
“Metaphor, Gesture and Thought” in
the transverse rather than the sagittal axis is dominant.35 On the
The Cambridge Handbook of Metaphor
basis of such evidence, the “right” direction for time could be
and Thought, ed. Raymond W. Gibbs
lateral, flowing left to right—but only in cultures where this orien-
(New York: Cambridge University Press,
tation maps to the reading direction.
2008), 492.
33 Geneviève Calbris, “From Left to Right…:
Coverbal Gestures and Their Symbolic
Direction in Graphical Representations of Time
Use of Space” in Metaphor and Gesture,
If we try to make a drawing of the sagittal timeline evoked by
vol. 3 of Gesture Studies, ed. Alan J.
verbal metaphor, the problem is an obvious one: The graphic
Cienki and Cornelia Müller (Amsterdam:
surface is normally orthogonal to our line of sight; we look
John Benjamins, 2008), 27-53.
more or less flat-on at a piece of paper, exhibition board, or digital
34 Alan J. Cienki, “Metaphoric Gestures and
display. Thus, a transformation is required to translate the sagittal
Some of their Relations to Verbal
Metaphorical Expressions” in Discourse
axis to one or another of the axes of the graphic surface. We rotate
and Cognition: Bridging the Gap, ed.
the sagittal z axis to the lateral x or vertical y. Alternatively, we
Jean-Pierre Koenig, (Stanford, CA:
try to present the sagittal line using some kind of perspectival
Center for the Study of Language and
depiction of an imaginary z axis that the user looks along. In
Information, 1998), 189-204.
physical media, this solution tends to be impractical; it has
35 Kensy Cooperrider and Rafael E.
Núñez, “Across Time, Across the
real y only started to come into its own with animated, interactive
Body: Transversal Temporal Gestures,”
digital representations.
Gesture 9 no. 2 (2009): 181–206 at 188.

When we look on a scene, the more distant parts of the
36 James J. Gibson, The Perception of the
observed world generally end up higher in our view than the
Visual World (Boston, MA: Houghton
nearer.36 Applying this perspective to the graphic surface suggests
Mifflin, 1950), 137-44.
DesignIssues: Volume 28, Number 4 Autumn 2012
9

---

Figure 5
a vertical mapping of the sagittal time axis, with distant time at the
A detail from Oresme’s Tractatus de
top and nearer time at the bottom. In dealing with the future, dis-
Configurationibus Qualitatum et Motuum
tant future events would be highest. However, in dealing with his-
(A Treatise on the Configurations of
tory, we are visualizing the past: Perhaps we imagine ourselves
Qualities and Motions). This copy in the
turning on the spot and looking behind. The recent past is now
British Library is from 1428, but the original
close and, in landscape terms, should be at the bottom of the view,
was probably made in the 1350s. The line
ab represents time, against which qualities
with the increasingly distant past higher up. It may be that less
and motions (i.e., quantitative variables)
cognitive load is imposed on the user by such a mapping to the
may be perpendicularly plotted. Nicole
vertical because in some respects it provides a kind of pictorial
Oresme, Tractatus de Configurationibus
view of the sagittal axis. Happily, this picture also accords with
Qualitatum et Motuum, cum prologo. 159-94.
Western textual conventions, in which earliest dates are listed first
© The British Library Board (Manuscript
Sloane 2156).
and continue down the page. Again, writing direction is formative.
In comparing three language cultures, Tversky, Kugelmass, and
Winter found strong evidence of its influence on the direction of
temporal diagram-making, time being more strongly influenced
by writing than were other variables.37
Past Struggles with the Graphic Orientation of Time
37 Barbara Tversky, Sol Kugelmass, and
A remarkable early example of drawing time is Nicole Oresme,
Atalia Winter, “Cross-Cultural and
whose work in the 1350s anticipates the quantitative time graphics
Developmental Trends in Graphic
Productions,” Cognitive Psychology 23
of the seventeenth and eighteenth centuries.38 His linear drawings
no. 4 (1991): 515-57.
of time are perhaps the first, though Oresme himself traces the
38 Marshall Clagett, ed. and transl., Nicole
idea to Aristotle (his Physics Book 4).39 Time is horizontal in all his
Oresme and the Medieval Geometry of
il ustrations, of which the most basic is a single line (see Figure 5).
Qualities and Motions: A Treatise on the
The influence of reading direction is emphasized by the use of
Uniformity and Difformity of Intensities
textual labels in alphabetic order. Nevertheless, Oresme too seems
Known as “Tractatus de configurationibus
qualitatum et motuum” (Madison, WI:
to have had trouble deciding on the “right” direction. He struggles
The University of Wisconsin Press, 1968).
to say whether the variable quantities mapped against time should
39 Marshall Clagett, The Science of
be referred to as latitudes or longitudes (Oresme trans. Clagett),40
Mechanics in the Middle Ages
opting for “latitudes” even though he then plots them vertical y,
(Madison, WI: University of Wisconsin
perpendicular to his horizontal time axis.
Press, 1959), 333.
40 Clagett,
Nicole Oresme and the Medieval

In looking at Barbeu-Dubourg’s and Priestley’s work, no
Geometry of Qualities and
one seems previously to have noticed that they, too, might have
Motions,169-73.
struggled with time’s direction. Barbeu-Dubourg’s chart uses the
10
DesignIssues: Volume 28, Number 4 Autumn 2012

---

Figure 6
A detail from Strass, F. 1849. Stream of Time,
or Chart of Universal History. 1849. [London]:
C. Smith, Mapseller. Note the struggle to
squeeze horizontal labels into vertically
oriented streams of time. Photo: Stephen
Boyd Davis. Collection: Stephen Boyd Davis.
horizontal axis for time and the vertical for categories, assigning
states and kingdoms to horizontal bands, with two additional
bands below them: one for events and one for individual historical
lives (again, see Figure 1). In French a suitable term for these divi-
sions would have been the word bande or, if emphasising its hori-
zontality, rang (row). Instead, we find the word colonne (column),
which in the eighteenth century could refer to printed columns of
text on a page.41 Dictionaries of the period give no hint that the
term was ever used for horizontal rows. His use of the word
strongly suggests that Barbeu-Dubourg originally intended to
have time running vertical y, with the categories as columns. The
term bande replaces colonne in the posthumous second edition.42
41 Dictionnaire de l’Académie Française, 4th
ed. (1762) http://artfl-project.uchicago.
Priestley’s explanation of his own chart shows a similar muddle:
edu/node/17 (accessed December 20,
The first edition has 32 occurrences of the word column.43 By the
2011).
seventh edition in 1778, only six occurrences remain, the others
42 Jacques Barbeu-Dubourg, Chronographie,
having been changed to division (20), space (4), or class (2).44 By 1800,
ou Description des Temps. (Chronography
the last few had been corrected, and the word column no longer
or Depiction of Time). (Paris, 1838).
British Library 798.1.19.
appears.45 This evidence surely suggests that Priestley original y
43 Priestley,
A Description of a Chart of
had in mind a vertical orientation for time.
Biography (1764).

We saw that the vertical orientation seems in some ways to
44 Joseph Priestley, A Description of a Chart
be the more direct or natural mapping of the sagittal axis—a
of Biography (London: Johnson, 1778).
potential y less burdensome intervention than rotation to the hori-
British Library: 10602.aa.22.
zontal. Why did both inventors change their minds? The most
45 Joseph Priestley, A Description of a Chart
of Biography (London: Johnson, 1800).
likely explanation is pragmatic: Names and labels are longer than
British Library: 10604.aaa.29.
they are tall, and fitting them into a vertical design is problematic.
DesignIssues: Volume 28, Number 4 Autumn 2012
11

---

Neither inventor could easily have reconciled his wish to label
every item with using a vertical time-axis. The vertical charts by
others who fol owed evade this problem by minimizing individual
lives and adding labels only to kingdoms and major events. Strass
in his Stream of Time46 does try to name every monarch within his
drawn tributaries, but the result is very small lettering, broken
lines and even broken words—and even then he is at times obliged
to rotate the names where the stream is particularly narrow (see
Figure 6). At both a pragmatic and a metaphorical level, then, the
direction of writing had a powerful influence on the earliest
Enlightenment timelines and, through them, on the historical
chronographics and statistical charts of others who fol owed.

To summarize before proceeding further, evidence from
psychology and anthropology suggests that time is almost univer-
sal y mapped metaphorical y to space. However, evidence from
language and from gesture points, quite literal y, in different direc-
tions. In addition, whether language or gesture is taken as the
model, evidence also suggests that reading direction has a power-
ful influence, both metaphorical y and pragmatical y. No single,
a-cultural, natural mapping for time exists, either in terms of ori-
entation or direction of travel. At the very least, this has implica-
tions for the design of chronographics intended for international or
multicultural audiences. From an exclusively Western beginning,
the textual architecture of the Web now pays significant attention
to scripting direction, character sets, ligatures, and other interna-
tional aspects of the word, but we see little evidence of such enter-
prise in relation to time.47 Research questions need to be pursued
about the relative cognitive burden of the competing orientations
and directions: Although we can make guesses based on the exist-
ing literature, such an approach is not real y good enough.
Af ordances of Timeline Orientations
46 Friedrich Strass, Stream of Time, or Chart
In addition to issues concerned with cognition, questions arise
of Universal History from the German of
concerning the differing effects of orientation on the observer. Do
Strass (New York: J.H. Colton, 1860)
the different orientations and directions carry messages? Perhaps
British Library: Maps 999.(74.)
the vertical and horizontal offer different affordances? Tversky et
47 See Elika J. Etemad, “Robust Vertical
Text Layout,” in 27th Internationalization
al. (1991) point out that in our embodied experience, the vertical
and Unicode Conference (Berlin, 2005),
dimension is asymmetric, with the down side literal y grounded
http://www.unicode.org/notes/tn22/
and the up side unbounded.48 In contrast, the horizontal, they
RobustVerticalLayout.pdf (accessed
argue, is symmetric. Therefore, thinking of horizontal time as
December 20, 2011), and Edward H.
potentially more neutral and less directed seems possible—
Trager, “International Text Layout and
although the examples of gestural metaphor already cited suggest
Typography: The Big and Future Picture,”
Gnome Live Text Layout Summit (Boston,
this neutrality is limited; in particular, loaded terms such as “sinis-
2006), http://unifont.org/textlayout/
ter” and “dexterity,” from the Latin for left and right, cannot be
TheBigPicture.pdf (accessed December
ignored. In the timelines I have studied, downward vertical orien-
20, 2011).
tation seems particularly popular for religious teleological repre-
48 Tversky, Kugelmass, and Winter, “Cross-
sentations. Perhaps the downward flow of time echoes the “Fall of
Cultural and Developmental Trends in
Graphic Productions,” 518.
12
DesignIssues: Volume 28, Number 4 Autumn 2012

---

Man” from an original state of grace in the Judaeo-Christian story.
For example, a vertical chart by Denny presents A Prophetical
Stream of Time flowing downward from God to The End, or Everlast-
ing State;49 meanwhile, Linton’s Chart of Ancient History similarly
traces the Biblical nations from the Creation downward.50 A glow-
ing orb at the top of Strass’ Stream of Time gives it a religious aura,
although the remainder of the content is secular: It, too, flows
downward. Upward flows of time are virtual y absent (but they
later reappear in another guise, as perspectival sagittal views).

Metaphors in early timelines included branches, chains,
streams, rivers, and arrows.51 One of the striking things about Bar-
beu-Dubourg and Priestley’s charts is the absence of explicit visual
rhetoric: They are both content to let the patterning of the data
speak for itself. In Tufte’s terms, there is no “chart junk.”52 Strass
objected to the very neutrality of Priestley’s view, favoring an
explicitly drawn grouping and linking of currents and tributaries
(see Figure 6). As an underlying metaphor, however, rivers are
49 Edward Denny, A chart, illustrated
important even to Priestley: In his explanatory Description, he
throughout with Pictorial Designs,
invokes them, citing their lack of beginning and end and likening
entitled, A Prophetical Stream of Time;
the lives of men to “so many small straws swimming on the sur-
or, an Outline of God’s Dealings with
Man, from the Creation to the End
face.”53 His chart bears the motto, Fluminis ritu feruntur, evoking
of all Things (Bath: Binns and Goodwin,
Horace’s advice to Maecenas to maintain his own position while
1849) British Library: Tab.539.c.
the world flows around him like a river. With the 1778 edition, he
50 Henry Linton. [no date]. Chart of Ancient
added a quotation to the Description, taken from Virgil’s Aeneid (I,
History. London: George Philip & Son.
118): Apparent rari nantes in gurgite vasto, “We see a few swimming
51 See, e.g., Daniel Rosenberg, “Joseph
Priestley and the Graphic Invention of
in the vast deep,” to describe the paucity of men of learning in the
Modern Time,” Studies in Eighteenth
Dark Ages.54 This apparently unitary river metaphor has two alter-
Century Culture 36 no. 1 (2007): 55–103;
native affordances. In the Description, we seem to stand on the riv-
and Davis, Bevan, and Kudikov, “Just
erbank watching time’s flow from outside—it flows past us
in Time.”
lateral y. Meanwhile, with the Horace quotation on the Chart itself,
52 Edward R. Tufte, The Visual Display of
Priestley puts the observer in the river as the waters swirl around
Quantitative Information (Cheshire, CT:
Graphics Press, 1983), 107-21.
him, presumably sagittally.
53 Priestley,
A Description of a Chart of

In relation to the first use, an obvious aspect of most of
Biography (1764), 24.
the chronographics discussed here is that they do indeed provide
54 Priestley,
A Description of a Chart of
an overview—a distancing perspective on time. We are familiar
Biography (1778), 23.
now with this idea of visualizations as summary overviews of
55 See Ben Shneiderman, “The Eyes Have It:
A Task by Data Type Taxonomy for
available data—an approach perhaps most strongly associated
Information Visualisations,” Proc. IEEE
with Shneiderman’s work at Maryland, repeatedly arguing the
Symposium on Visual Languages
value of seeing entire datasets within a single view.55 This benefit
(Boulder, CO: IEEE Computer Society,
was familiar to Priestley, whose argument is peppered with
Washington DC, 1996), 336-43; Also see
phrases such as “at a single glance,” “at one glance,” “the noblest
Stuart K. Card, Jock D. Mackinlay, and
prospect,” “a comprehensive view of the succession of great men,”
Ben Shneiderman, eds., Readings in
Information Visualization: Using Vision to
and “the contents of the tablet will be fixed in the imagination, not
Think (San Francisco: Morgan Kaufmann,
in succession, but at once.”56
1999), 285-305. Chapter 3.3 Overview

A less obvious potentiality is to create a sense of immersion
and Detail.
in the historic moment—a metaphor surely more strongly tied to
56 Joseph Priestley, A Description of a
the sagittal view. We can glimpse this in Priestley’s description,
Chart of Biography (1764), 9, 10, 25, 3,
and 7, respectively.
DesignIssues: Volume 28, Number 4 Autumn 2012
13

---

Figure 7
already quoted, of Sir Isaac Newton in the circle of his friends,
BBC timeline, A History of the World,
but the greater champion of immersion is Barbeu-Dubourg. He
Version 1.1 2010. The z-axis, orthogonal to
inserted his 16.5-meter continuous paper roll of time into a machine
the picture plane, is used for time. In this
chronographique, with small crank handles to wind history back
case, the designers have chosen to put the
and forth: About 150 years are visible at any moment. The machine
present time in the far distance, in front of
the observer. Behind the observer is the
was enthusiastically described in great detail by Diderot in
most distant past. BBC Radio 4, A History
the Encyclopédie.57 Meanwhile, Barbeu-Dubourg’s own pamphlet
of the World http://www.bbc.co.uk/
evokes “a moving, living tableau, through which pass in review al
ahistoryoftheworld (used with permission).
the ages of the world […] where the rise and fall of Empires are
acted out in visible form. . ”58 Here, the sense of history as immer-
sive experience is uppermost. We should remember that Barbeu-
Dubourg almost certainly intended at first that time should flow
vertically in his machine. Perhaps vertical orientations afford
a stronger sense of immersion than the horizontal, with the
more intrusive rotation to the horizontal producing a distancing,
objectifying effect. This possibility raises research questions
concerned with af ect, adding to those questions already raised in
relation to cognition.
57 Denis Diderot, “Chronologique

Instead of mapping the sagittal to the vertical or horizontal,
(machine),” Encyclopédie ou Dictionnaire
the designer can create a perspectival view of time, looking along
Raisonné des Sciences, des Arts et des
the sagittal z axis. Here, the immersive aspect is one of the most
Metiers vol. 3 (400-01), trans. Stephen
Boyd Davis, The Encyclopedia of Diderot
appealing features: We seem to stare directly through time from a
& d’Alembert Collaborative Translation
point within it. Such views have become popular in digital media.
Project (Ann Arbor: University of
They tend to sacrifice some of the benefits of overview, not least
Michigan, 2009), http://hdl.handle.
because as users interact, much of the timeline passes out of view
net/2027/spo.did2222.0001.081
“behind” them. More distant items may be too small to read. The
(accessed December 20, 2011).
58 Barbeu-Dubourg,
Chronographie, ou
problems of which way time should flow and quite how it should
Description des Tems (Chronography, or
be rendered remain. In a 2010 timeline of historical artifacts for the
Depiction of Time), 8.
14
DesignIssues: Volume 28, Number 4 Autumn 2012

---

Figure 8
BBC (see Figure 7), present time is in the far distance, in front of
Guardian online newspaper timeline of the
the user, while “behind” the user and out of view is the most dis-
“Arab Spring” 2011. Again, the future is
tant past. Clearly this affords a novel view on history: It is unusual
ahead of the user, but here time is presented
to look at the present from the past. A similar approach, also look-
as an incline, a hybrid of sagittal and vertical
ing toward the future, is offered by recent timelines for the Guard-
views. The curvature of the timeline surface
helps avoid problems of scale. © Guardian
ian newspaper website (see Figure 8); but instead of looking into a
News & Media Ltd. 2011.
kind of time tunnel, we now look at time as a curved incline—a
hybrid of sagittal and vertical views. This curve has some particu-
lar advantages in terms of scale. Instead of distant items becoming
il egibly small, the upwardly curved, distant portion of the time-
line ensures that they remain visible further into the distance.
Note that the perspectival effect al ows us to accept the otherwise
59 Robin L. Kullberg, Dynamic Timelines:
“unnatural” idea that later times are “higher” than earlier ones;
Visualizing Historical Information in Three
whereas, when time is diagrammatic rather than pictorial, this
Dimensions, MSc Dissertation, MIT,
view is not an acceptable option.
September 1995. http://citeseerx.ist.psu.
edu/viewdoc/download?doi=10.1.1.51.52

Extrapolating beyond the use of simple, three-dimensional
78&rep=rep1&type=pdf (accessed
interfaces, the temptation is to look for advantages in virtual envi-
December 21, 2011).
ronments. In a 3D timeline of photography by Kul berg, individual
60 See, e.g., Nigel Foreman, Stephen Boyd
lifelines are laid out on a virtual surface, while individual photo-
Davis, Magnus Moar, Liliya Korallo,
graphs stand on these lifelines at the appropriate points in
and Emma Chappell, “Can Virtual
time.59 Events from contextual history appear on a lower, darker
Environments Enhance the Learning of
Historical Chronology?” Instructional
layer. The user can look at time from many points of view and dis-
Science 36 no. 2 (2008): 155-73; and
tances. Virtual environment representations for time have been
Liliya Korallo, Nigel Foreman, Stephen
compared against others experimental y.60 Unfortunately, for some
Boyd Davis, Magnus Moar, and Mark
groups of users, a simple horizontal array proved more effective in
Coulson, “Can Multiple ‘Spatial’ Virtual
terms of recall than the virtual 3D space. Further work is needed to
Timelines Convey the Relatedness of
Chronological Knowledge Across Parallel
discover the determinants of successful design in this aspect of
Domains?” Computers & Education 58
chronographics, as in so many others.
no. 2 (2012): 856-62.
DesignIssues: Volume 28, Number 4 Autumn 2012
15

---

Note also how two related issues also impinge strongly on
our perception of graphic time: scale and calibration. Barbeu-
Dubourg and Priestley both strongly advocated linear mappings,
assuming that this presents a true impression of time. A tricky
epistemological problem arises here: Evidence suggests that the
mental models of time we routinely use are non-linear.61 Drucker
and Nowviskie have argued that digital tools should capture the
subjectivity of time perception in “elaborate, subjectively inflected
timelines.”62 Continuum, a project at University of Southampton,
demonstrates the benefits of suppressing whole chunks of time so
that separate historical periods can be juxtaposed.63 It also offers
multiple views of time on screen at the same moment. Strong argu-
ments can be made both for linear and non-linear views, depend-
61 See Paul Fraisse, “Perception and
ing on how we decide time “really is” and what purposes the
Estimation of Time,” Annual Review of
Psychology 35 (1984): 1-37; Gal
representation serves.
Zauberman, B. Kyu Kim, Selin A. Malkoc,

If western writing direction has often overridden other
and James R. Bettman, “Discounting
models of time’s flow, this same dominance affects the units and
Time and Time Discounting: Subjective
measures of the timeline; nevertheless, there are several reasons
Time Perception and Intertemporal
why differing measures of time may be needed. One is to reflect
Preferences,” Journal of Market
Research 45 (2008): 543–56; and Eviatar
the varied cultures of the users. Dershowitz and Reingold list 30
Zerubavel, Time Maps: Collective
different calendars, including several in current use today and
Memory and the Social Shape of the Past
many that are significant in history.64 Within a single culture, rival
(Chicago: University of Chicago Press,
dating schemes may be used because of differing scholarly opin-
2003).
ion. In addition, having more than one calibration of a chart simply
62 Johanna Drucker and Bethany Nowviskie,
may be useful—for example, dates counted forward from a point
“Speculative Computing: Aesthetic
Provocations in Humanities Computing,”
in history to our own time, as well as dates counted backward
in A Companion to Digital Humanities,
from the present day. The use of multiple and alternative calendars
ed. Susan Schreibman, Ray Siemens, and
is culturally inclusive, and it is a necessity when dealing with
John Unsworth (Oxford: Blackwell, 2004)
many kinds of history.
442-43.
63 Paul André, Max L. Wilson, Alistair
Improving the Design of Timelines—and a Research Agenda
Russell, Daniel A. Smith, Alisdair Owens,
and m.c. schraefel [sic], “Continuum:
Timelines undoubtedly will continue to be seen as useful, and the
Designing Timelines for Hierarchies,
enhanced capabilities of digital media, in particular for the Web,
Relationships and Scale” in Proc. UIST
offer increased roles for design. As new chronographic designs
(New York: ACM, 2007), 101-10.
emerge, new discoveries will be made about data. Manovich uses
64 Nachum Dershowitz and Edward M.
cultural analytics to plot cultural artifacts against time in search of
Reingold,Calendrical Calculations
3rd ed. (New York: Cambridge University
new insights.65 The Historical Interactive Timeline (HiT) al ows users
Press, 2008).
to create complex mixed queries to see patterns in a dataset.66 What
65 Lev Manovich, “What Is Visualization?”
new knowledge can be created by advanced forms of chrono-
in Lev Manovich: Cultural Analytics,
graphic design? Interactivity enables users to make their own
Software Studies, New Media,
choices—but, only if the design supports them. The designer must
Digital Humanities. http://manovich.
create flexible solutions that adapt to changing content and con-
net/2010/10/25/new-article-what-is-
visualization/ (accessed December 21,
text, rather than specifying appearance to the last pixel. Although
2011).
flexible tools for constructing timelines, such as SIMILE, are
66 Davis, Bevan, and Kudikov, “Just in
becoming more sophisticated, they have far to go.67 Rather than
Time,” (accessed December 17, 2011).
assuming that designing timelines and other chronographics is
67 SIMILE Timeline, Massachusetts Institute
unproblematic, we should embrace the often difficult implications
of Technology. www.simile-widgets.org/
timeline/ (accessed December 21, 2011).
for the designer and the researcher. Which are the most effective
16
DesignIssues: Volume 28, Number 4 Autumn 2012

---

models and viewpoints for various users and purposes? Does
looking onto time “from the side” have a greater objectifying effect
than looking along time? How can problems of occlusion and data
density be solved in views that look “along” time? What are the
criteria for choosing from among 2D, 2½D, and 3D models? What
are the cognitive and affective consequences of different locations
of user viewpoint? Which are the best ways of mapping time for
particular purposes? Does the user benefit from multiple synchro-
nized views? Is a single zoomable view preferable? How can visual
constancy be maintained between views? What use can be made of
“intel igent” levels of detail? What about non-linear views, such as
fisheye lenses? Can we evaluate experimental y the advantages
claimed by Priestley and Shneiderman more than two centuries
apart for showing entire datasets in a single view?

In answering such questions, we are not without help from
existing research in many fields, but the problem remains of apply-
ing this knowledge to the issues I have raised. Exciting opportuni-
ties for new research exist, while those interested in these issues
can also grapple with them by exploring through designing.
Acknowledgements
I am very grateful to Professor Michael Twyman of Reading Uni-
versity for al owing me access to his personal col ection of chrono-
graphic materials, to Professor Anthony Grafton at Princeton for
valuable advice, to the staff at Chetham’s Library Manchester and
at Princeton University Library especial y Stephen Ferguson, Cura-
tor of Rare Books, and to Christine North for advice on translation.

DesignIssues: Volume 28, Number 4 Autumn 2012
17

---