SlideShare a Scribd company logo
Twyman, M (1981)

‘Typography without words’,

in Visible Language, XV no. 1, 5–12
Spencer, Herbert (1969)

The visible word

London: Lund Humphries/Royal College of Art
Typographic hierarchy
Legibility research
Twyman 1981
A notation system to guide discussion and design decisions
Notation that describes function rather than exact form
(making use of the technical limitations of a typewriter)
Spatial grouping (Gestalt psychology): vertical and lateral
spatial cues
To show the typography is a modular system
About the relationship between levels of heading and
paragraphs of continuous text
Which examples would be ‘reasonably unambiguous in visual terms’?
A
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
ooooo
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
ooooo
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
D
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
ooooo
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
ooooo
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
B
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
ooooo
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
ooooo
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
E
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
ooooo
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
ooooo
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
C
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
ooooo
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
oooo xxxxxxxxxxxxxxxxxxxxx
ooo xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
F
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
oooo xxxxxxxxxxxxxxxxxxxxx
ooo xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
ooooo
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
Which examples would be ‘reasonably unambiguous in visual terms’?
largest ethnic group in East Asia. Other languages
also use variant forms of Hanzi: Kanji (Japanese),
Hanja (Korean), and Chữ Nôm (Vietnamese).
The Chinese language
There are eight main varieties of speech in China
thataredistinctlydifferentfromeachotherinterms
of pronunciation and vocabulary, though they are
often referred to as ‘dialects’ rather than languages.
Linguistic aspects of Hanzi
Chinese characters are classified as logographic,
meaning that each character represents a mor-
pheme, the smallest unit of meaning. It is often
said that Chinese characters are pictographic; this
is only partly true.
largest ethnic group in East Asia. Other languages
also use variant forms of Hanzi: Kanji (Japanese),
Hanja (Korean), and Chữ Nôm (Vietnamese).
The Chinese language
There are eight main varieties of speech in China
thataredistinctlydifferentfromeachotherinterms
of pronunciation and vocabulary, though they are
often referred to as ‘dialects’ rather than languages.
Linguistic aspects of Hanzi
Chinese characters are classified as logographic,
meaning that each character represents a mor-
pheme, the smallest unit of meaning. It is often
said that Chinese characters are pictographic; this
is only partly true.
largest ethnic group in East Asia. Other languages
also use variant forms of Hanzi: Kanji (Japanese),
Hanja (Korean), and Chữ Nôm (Vietnamese).
The Chinese language
There are eight main varieties of speech in China
thataredistinctlydifferentfromeachotherinterms
of pronunciation and vocabulary, though they are
often referred to as ‘dialects’ rather than languages.
Linguistic aspects of Hanzi
Chinese characters are classified as logographic,
meaning that each character represents a mor-
pheme, the smallest unit of meaning. It is often
said that Chinese characters are pictographic; this
is only partly true.
largest ethnic group in East Asia. Other languages
also use variant forms of Hanzi: Kanji (Japanese),
Hanja (Korean), and Chữ Nôm (Vietnamese).
The Chinese language
There are eight main varieties of speech in China
thataredistinctlydifferentfromeachotherinterms
of pronunciation and vocabulary, though they are
often referred to as ‘dialects’ rather than languages.
Linguistic aspects of Hanzi
Chinese characters are classified as logographic,
meaning that each character represents a mor-
pheme, the smallest unit of meaning. It is often
said that Chinese characters are pictographic; this
is only partly true.
A B
D E
✔
✔
Which examples would be ‘reasonably unambiguous in visual terms’?
largest ethnic group in East Asia. Other languages
also use variant forms of Hanzi: Kanji (Japanese),
Hanja (Korean), and Chữ Nôm (Vietnamese).
The Chinese language
There are eight main varieties of speech in China
thataredistinctlydifferentfromeachotherinterms
of pronunciation and vocabulary, though they are
often referred to as ‘dialects’ rather than languages.
Chinese characters are classified as logographic,
meaning that each character represents a mor-
pheme, the smallest unit of meaning. It is often
said that Chinese characters are pictographic; this
is only partly true.
Linguistic
aspects of Hanzi
largest ethnic group in East Asia. Other languages
also use variant forms of Hanzi: Kanji (Japanese),
Hanja (Korean), and Chữ Nôm (Vietnamese).
There are eight main varieties of speech in China
thataredistinctlydifferentfromeachotherinterms
of pronunciation and vocabulary, though they are
often referred to as ‘dialects’ rather than languages.
Linguistic aspects of Hanzi
Chinese characters are classified as logographic,
meaning that each character represents a mor-
pheme, the smallest unit of meaning. It is often
said that Chinese characters are pictographic; this
is only partly true.
The Chinese
language
C
F
Which example articulate the text as clearly as possible?
A
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
ooooo
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
B
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
ooooo
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
C
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
ooooo
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
D
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
ooooo
xxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
Which example articulate the text as clearly as possible?
Continuous prose in Chinese is customarily set
justified left and right. This is true even in clas-
sical poetry, where phrases have set number of
characters. The fact that Chinese characters are
monowidth makes flush left, ragged right align-
ment impractical.
Spacing and layout
As mentioned earlier, spacing in Chinese typefaces
is achieved internally within each character, by
way of defining a surface frame that is a smaller
percentage than the body frame (em square). This
internal spacing is the optimal spacing under most
circumstances.Columnwidths(hencelinelengths)
are calculated by number of characters.
Since Chinese is monowidth, and that column
widths are measured by fixed numbers of charac-
ters, spacing should be an easy matter. But this
is not the case. There are three factors that make
justification a complex issue: line-beginning and
line-ending rules; the insertion of proportional
width characters, such as Latin glyphs, symbols
or figures; adjustments of spacing around certain
punctuation marks.
Continuous prose in Chinese is customarily set
justified left and right. This is true even in clas-
sical poetry, where phrases have set number of
characters. The fact that Chinese characters are
monowidth makes flush left, ragged right align-
ment impractical.
Spacing and layout
As mentioned earlier, spacing in Chinese typefaces
is achieved internally within each character, by
way of defining a surface frame that is a smaller
percentage than the body frame (em square). This
internal spacing is the optimal spacing under most
circumstances.Columnwidths(hencelinelengths)
are calculated by number of characters.
Since Chinese is monowidth, and that column
widths are measured by fixed numbers of char-
acters, spacing should be an easy matter. But this
is not the case. There are three factors that make
justification a complex issue: line-beginning and
line-ending rules; the insertion of proportional
width characters, such as Latin glyphs, symbols
or figures; adjustments of spacing around certain
punctuation marks.
A B
Which example articulate the text as clearly as possible?
Continuous prose in Chinese is customarily set
justified left and right. This is true even in clas-
sical poetry, where phrases have set number of
characters. The fact that Chinese characters are
monowidth makes flush left, ragged right align-
ment impractical.
Spacing and layout
As mentioned earlier, spacing in Chinese typefaces
is achieved internally within each character, by
way of defining a surface frame that is a smaller
percentage than the body frame (em square). This
internal spacing is the optimal spacing under most
circumstances.Columnwidths(hencelinelengths)
are calculated by number of characters.
Since Chinese is monowidth, and that column
widths are measured by fixed numbers of charac-
ters, spacing should be an easy matter. But this
is not the case. There are three factors that make
justification a complex issue: line-beginning and
line-ending rules; the insertion of proportional
width characters, such as Latin glyphs, symbols
or figures; adjustments of spacing around certain
punctuation marks.
Continuous prose in Chinese is customarily set
justified left and right. This is true even in clas-
sical poetry, where phrases have set number of
characters. The fact that Chinese characters are
monowidth makes flush left, ragged right align-
ment impractical.
Spacing and layout
As mentioned earlier, spacing in Chinese typefaces
is achieved internally within each character, by
way of defining a surface frame that is a smaller
percentage than the body frame (em square). This
internal spacing is the optimal spacing under most
circumstances.Columnwidths(hencelinelengths)
are calculated by number of characters.
Since Chinese is monowidth, and that column
widths are measured by fixed numbers of charac-
ters, spacing should be an easy matter. But this
is not the case. There are three factors that make
justification a complex issue: line-beginning and
line-ending rules; the insertion of proportional
width characters, such as Latin glyphs, symbols
or figures; adjustments of spacing around certain
C D
✔
moulds to mass produce individual types, which is an essential criteria for
the development of typefaces.
Chinese (Hanzi)
Chinese is currently the most widely used language in the world, used by
1302 million people (Lewis et al 2016). The Chinese language is written
in Hanzi, where Han refers to the dynasty (202 BC–220 AD) as well as the
largest ethnic group in East Asia. Other languages also use variant forms
of Hanzi: Kanji (Japanese), Hanja (Korean), and Chữ Nôm (Vietnamese).
The Chinese language and the Hanzi script
There are eight main varieties of speech in China that are distinctly dif-
ferent from each other in terms of pronunciation and vocabulary, though
they are often referred to as ‘dialects’ rather than languages.
Linguistic aspects of Hanzi
Chinese characters are classified as logographic, meaning that each char-
acter represents a morpheme, the smallest unit of meaning. It is often said
that Chinese characters are pictographic; this is only partly true.
Traditional and Simplified Chinese There are currently
two forms of Chinese orthography: Traditional Chinese and Simplified
Chinese. Simplified Chinese characters were derived from the traditional
forms by reducing the number of strokes, and were first proposed during
the Republic of China era in 1935.
Anatomical characteristics of Chinese characters
Chinese characters are customarily monowidth; they sit within a notional
square. In typesetting, this notional square is defined as the em square,
How many orders of heading?
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
ooooo
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
ooooo
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
iiii
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
iiii xxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxx
ooooo
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
1
2
3
4
is therefore important that justification is done via Chinese justification
rules rather than Latin ones to ensure good spacing.
Code-mixing
Code-mixing is a term borrowed from linguistics. It is referred to the in-
sertion of words and short phrases from another script into the text flow
of the main script.
Explanation or disambiguation
Chinese translations of technical or specialised terms may be ambiguous
because they are little known or non-standard. This may also be true for
names of places and foreign people.
Parallel typesetting
Text extent
A passage translated from English into Chinese is usually shorter. This is
due to the fact that Chinese is a more concise writing system.
Typesize
The anatomical differences between Chinese and Latin characters make
them appear different in size even when they are both set in the same
point size. Chinese characters appear larger, as they occupy the em square
more fully than Latin characters.
Weight and density Chinese characters can have anywhere
between one to more than 30 strokes, making their densities vary quite
considerably. A page of Chinese text can therefore appear rather spotty
Which is the second order of heading?
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
ooooo
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
iiii
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
ooooo
ooooo
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
iiii
xxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxx
iiii xxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxx
Content structure
The norm: the main text
elements above
the norm
elements below
the norm
eg chapter headings

subheads

images

introductory text

pull quotes

etc . . .
eg footnotes

captions

margin notes

page numbers

headers and footers

etc . . .
continuous text

body text
Typographic variations for headings: differentiation from the norm
Continuous text: roman
Heading: italic
Heading: small caps
HEADING: ALL CAPS
Heading: bold
Heading: another typeface
Heading: colour
Heading: size
Heading differentiation: line break and space above
each glyph, with a slightly smaller square defining the margin within which
the character sits, referred to as the ‘surface frame’ (Lu and Tang 2016:111).
Typeface designers define surface frames as percentages of the body frame.
The larger this percentage, the more tightly spaced the typeface appear.
Units of measurement
The Western point system was brought into China along with the in-
dustrialised process of printing with moveable lead type in 1859–60 by
Presbyterian missionary William Gamble. A modular size system with nu-
merical labels were subsequently adopted for Chinese lead type, with nine
related sizes which were mapped onto points. The most common size for
continuous text was number 5, equal to 10.5 point. Latin characters were
casted in this size for the purpose of code-mixed setting. Chinese digi-
tal typesetting now adopts the Western system of points and picas. The
concept of the em is important in Chinese typesetting, as column widths,
spacing and indents are defined using fractions or multiples of the em,
which is the point size of the current type used.
Heading differentiation: equal space above and below (not ideal)
each glyph, with a slightly smaller square defining the margin within which
the character sits, referred to as the ‘surface frame’ (Lu and Tang 2016:111).
Typeface designers define surface frames as percentages of the body frame.
The larger this percentage, the more tightly spaced the typeface appear.
Units of measurement
The Western point system was brought into China along with the in-
dustrialised process of printing with moveable lead type in 1859–60 by
Presbyterian missionary William Gamble. A modular size system with nu-
merical labels were subsequently adopted for Chinese lead type, with nine
related sizes which were mapped onto points. The most common size for
continuous text was number 5, equal to 10.5 point. Latin characters were
casted in this size for the purpose of code-mixed setting. Chinese digi-
tal typesetting now adopts the Western system of points and picas. The
concept of the em is important in Chinese typesetting, as column widths,
spacing and indents are defined using fractions or multiples of the em,
which is the point size of the current type used.
Heading differentiation: closer to paragraph below than above
each glyph, with a slightly smaller square defining the margin within which
the character sits, referred to as the ‘surface frame’ (Lu and Tang 2016:111).
Typeface designers define surface frames as percentages of the body frame.
The larger this percentage, the more tightly spaced the typeface appear.
Units of measurement
The Western point system was brought into China along with the in-
dustrialised process of printing with moveable lead type in 1859–60 by
Presbyterian missionary William Gamble. A modular size system with nu-
merical labels were subsequently adopted for Chinese lead type, with nine
related sizes which were mapped onto points. The most common size for
continuous text was number 5, equal to 10.5 point. Latin characters were
casted in this size for the purpose of code-mixed setting. Chinese digi-
tal typesetting now adopts the Western system of points and picas. The
concept of the em is important in Chinese typesetting, as column widths,
spacing and indents are defined using fractions or multiples of the em,
which is the point size of the current type used.
Heading differentiation: now bold
each glyph, with a slightly smaller square defining the margin within which
the character sits, referred to as the ‘surface frame’ (Lu and Tang 2016:111).
Typeface designers define surface frames as percentages of the body frame.
The larger this percentage, the more tightly spaced the typeface appear.
Units of measurement
The Western point system was brought into China along with the in-
dustrialised process of printing with moveable lead type in 1859–60 by
Presbyterian missionary William Gamble. A modular size system with nu-
merical labels were subsequently adopted for Chinese lead type, with nine
related sizes which were mapped onto points. The most common size for
continuous text was number 5, equal to 10.5 point. Latin characters were
casted in this size for the purpose of code-mixed setting. Chinese digi-
tal typesetting now adopts the Western system of points and picas. The
concept of the em is important in Chinese typesetting, as column widths,
spacing and indents are defined using fractions or multiples of the em,
which is the point size of the current type used.
Heading differentiation: bold and left indent
each glyph, with a slightly smaller square defining the margin within which
the character sits, referred to as the ‘surface frame’ (Lu and Tang 2016:111).
Typeface designers define surface frames as percentages of the body frame.
The larger this percentage, the more tightly spaced the typeface appear.
Units of measurement
The Western point system was brought into China along with the in-
dustrialised process of printing with moveable lead type in 1859–60 by
Presbyterian missionary William Gamble. A modular size system with nu-
merical labels were subsequently adopted for Chinese lead type, with nine
related sizes which were mapped onto points. The most common size for
continuous text was number 5, equal to 10.5 point. Latin characters were
casted in this size for the purpose of code-mixed setting. Chinese digi-
tal typesetting now adopts the Western system of points and picas. The
concept of the em is important in Chinese typesetting, as column widths,
spacing and indents are defined using fractions or multiples of the em,
which is the point size of the current type used.
Heading differentiation: larger and left indent
each glyph, with a slightly smaller square defining the margin within which
the character sits, referred to as the ‘surface frame’ (Lu and Tang 2016:111).
Typeface designers define surface frames as percentages of the body frame.
The larger this percentage, the more tightly spaced the typeface appear.
Units of measurement
The Western point system was brought into China along with the in-
dustrialised process of printing with moveable lead type in 1859–60 by
Presbyterian missionary William Gamble. A modular size system with nu-
merical labels were subsequently adopted for Chinese lead type, with nine
related sizes which were mapped onto points. The most common size for
continuous text was number 5, equal to 10.5 point. Latin characters were
casted in this size for the purpose of code-mixed setting. Chinese digi-
tal typesetting now adopts the Western system of points and picas. The
concept of the em is important in Chinese typesetting, as column widths,
spacing and indents are defined using fractions or multiples of the em,
which is the point size of the current type used.
Heading differentiation: larger and left indent (smaller line spaces)
each glyph, with a slightly smaller square defining the margin within which
the character sits, referred to as the ‘surface frame’ (Lu and Tang 2016:111).
Typeface designers define surface frames as percentages of the body frame.
The larger this percentage, the more tightly spaced the typeface appear.
Units of measurement
The Western point system was brought into China along with the in-
dustrialised process of printing with moveable lead type in 1859–60 by
Presbyterian missionary William Gamble. A modular size system with nu-
merical labels were subsequently adopted for Chinese lead type, with nine
related sizes which were mapped onto points. The most common size for
continuous text was number 5, equal to 10.5 point. Latin characters were
casted in this size for the purpose of code-mixed setting. Chinese digi-
tal typesetting now adopts the Western system of points and picas. The
concept of the em is important in Chinese typesetting, as column widths,
spacing and indents are defined using fractions or multiples of the em,
which is the point size of the current type used.
Differentiation of heading: larger and outdent
each glyph, with a slightly smaller square defining the margin within which
the character sits, referred to as the ‘surface frame’ (Lu and Tang 2016:111).
Typeface designers define surface frames as percentages of the body frame.
The larger this percentage, the more tightly spaced the typeface appear.
Units of measurement
The Western point system was brought into China along with the in-
dustrialised process of printing with moveable lead type in 1859–60 by
Presbyterian missionary William Gamble. A modular size system with nu-
merical labels were subsequently adopted for Chinese lead type, with nine
related sizes which were mapped onto points. The most common size for
continuous text was number 5, equal to 10.5 point. Latin characters were
casted in this size for the purpose of code-mixed setting. Chinese digi-
tal typesetting now adopts the Western system of points and picas. The
concept of the em is important in Chinese typesetting, as column widths,
spacing and indents are defined using fractions or multiples of the em,
which is the point size of the current type used.
may occur when words run together. This should be avoided when writing
and editing the text. Punctuation and typographic treatments can also be
used to avoid ambiguity.
Anatomical characteristics of Chinese characters
Chinese characters are customarily monowidth; they sit within a notional
square. In typesetting, this notional square is defined as the em square,
referred to as the ‘body frame’. Spacing between Chinese characters are
defined internally within each glyph, with a slightly smaller square defin-
ing the margin within which the character sits, referred to as the ‘surface
frame’ (Lu and Tang 2016:111). Typeface designers define surface frames as
percentages of the body frame. The larger this percentage, the more tightly
spaced the typeface appear.
Units of measurement
The Western point system was brought into China along with the in-
dustrialised process of printing with moveable lead type in 1859–60 by
Presbyterian missionary William Gamble. A modular size system with nu-
merical labels were subsequently adopted for Chinese lead type, with nine
related sizes which were mapped onto points. The most common size for
[paragraph]
[heading 1]
[paragraph]
[heading 2]
Spencer 1969
Review of literature on legibility research of over 100 years
Empirical studies conducted by researchers outside the
typography discipline
Controlled experiments in lab situations
Studies that guided typographic design
Spencer 1969
Distinctiveness of characters
Legibility of typefaces
Importance of x-height rather than point size
Capitals, lowercase, bold, italics, numerals, punctuation
Interrelationship of type size, measure and leading
Indication of paragraphs and alignment (line endings)
Congeniality, atmosphere value and typographic allusion
Legibility research: methods
Distance
Speed of perception (tachistoscope)
Eye-movement
Blink-rate
Visual fatigue
Peripheral vision
Focal variation
Visibility
Heart-rate measurement
Haploscope (stereoscope)
Shaking table
Rate of work
We read in series of jumps along a line
Javal 1878
The quick brown fox jumps over the lazy dog. Five boxing gymnasts jump
fixations (pauses): where perception occurs
94% of reading time devoted to pauses
each pause ¼ second
saccades (jumps)
x-height is as important if not more important than size for legibility
Poulton 1955
Handgloves
HandglovesGill Sans: smaller x-height
Helvetica: larger x-height
x-height is as important if not more important than size for legibility
Poulton 1955
Gill Sans 170pt
Helvetica 150pt
Handgloves
Handgloves
x-heights have been equalised
Lead type
https://upload.wikimedia.org/wikipedia/commons/5/5a/Plantin_letterpress.jpg
Serif and sans serif typefaces equally preferred by readers and read
Hartley & Room 1983, Tinker 1963, Gould et al 1987, Zachrisson 1965
Legibility research in printing is
concerned with the efficiency of the
visible word. So, too is the practice
of typographical design. During the
past century both researchers and
designers have put forward
proposals for making printed letters
communicate more efficiently. This
report describes and illustrates some
of the more significant of these
proposals.
Legibility research in printing is
concerned with the efficiency of the
visible word. So, too is the practice
of typographical design. During the
past century both researchers and
designers have put forward
proposals for making printed letters
communicate more efficiently. This
report describes and illustrates
some of the more significant of
these proposals.
Serif faces may be easier to read in continuous text than sans serif faces
Burt 1959, Hvistendal & Kahl 1975, Robinson, Abbamonte & Evans 1971,Wheeldon 1995
All capitals setting slows down reading speed by 13–20%
Breland & Breland 1944
Legibility research in printing is
concerned with the efficiency of the
visible word. So, too is the practice
of typographical design. During the
past century both researchers and
designers have put forward
proposals for making printed letters
communicate more efficiently. This
report describes and illustrates some
of the more significant of these
proposals.
LEGIBILITY RESEARCH IN
PRINTING IS CONCERNED
WITH THE EFFICIENCY OF THE
VISIBLE WORD. SO, TOO IS THE
PRACTICE OF TYPOGRAPHICAL
DESIGN. DURING THE PAST
CENTURY BOTH RESEARCHERS
AND DESIGNERS HAVE PUT
FORWARD PROPOSALS FOR
MAKING PRINTED LETTERS
COMMUNICATE MORE
EFFICIENTLY. THIS REPORT
DESCRIBES AND ILLUSTRATES
SOME OF THE MORE
SIGNIFICANT OF THESE
PROPOSALS.
Word shapes are more distinctive in upper- and lowercase setting
Ireland 1944; Paterson and Tinker 1928, 1940; Starch 1914
Handgloves
HANDGLOVES
Top halves of lowercase letters easier to recognise than lower halves
Messmer 1903
Handgloves
Handgloves
Italic in continuous text can reduce reading speed substantially
Foster & Bruce 1982, Tinker 1955
Legibility research in printing is
concerned with the efficiency of the
visible word. So, too is the practice
of typographical design. During the
past century both researchers and
designers have put forward
proposals for making printed letters
communicate more efficiently. This
report describes and illustrates some
of the more significant of these
proposals.
Legibility research in printing is
concerned with the efficiency of the
visible word. So, too is the practice of
typographical design. During the past
century both researchers and
designers have put forward proposals
for making printed letters
communicate more efficiently. This
report describes and illustrates some
of the more significant of these
proposals.
Readers preferred faces ‘approaching the appearance of bold face type’
Paterson and Tinker 1940
Legibility research in printing is
concerned with the efficiency of the
visible word. So, too is the practice
of typographical design. During the
past century both researchers and
designers have put forward
proposals for making printed letters
communicate more efficiently. This
report describes and illustrates some
of the more significant of these
proposals.
Legibility research in printing is
concerned with the efficiency of the
visible word. So, too is the practice
of typographical design. During the
past century both researchers and
designers have put forward
proposals for making printed
letters communicate more
efficiently. This report describes
and illustrates some of the more
significant of these proposals.
1–4 points of leading read faster than set solid (no leading)
Becker, Heinrich, van Sichowsny & Wendt 1970, Bentley 1921, Paterson & Tinker 1932
Legibility research in printing is
concerned with the efficiency of the
visible word. So, too is the practice
of typographical design. During the
past century both researchers and
designers have put forward
proposals for making printed letters
communicate more efficiently. This
report describes and illustrates some
of the more significant of these
proposals.
Legibility research in printing is
concerned with the efficiency of the
visible word. So, too is the practice
of typographical design. During the
past century both researchers and
designers have put forward
proposals for making printed letters
communicate more efficiently. This
report describes and illustrates some
of the more significant of these
proposals.
Lower contrast oblique axis (Minion) / high contrast vertical axis (Didot)
Minion (left), Didot (right) / Wiggins 1967 (Anisson 1790s)
Legibility research in printing is
concerned with the efficiency of the
visible word. So, too is the practice
of typographical design. During the
past century both researchers and
designers have put forward
proposals for making printed letters
communicate more efficiently. This
report describes and illustrates some
of the more significant of these
proposals.
Legibility research in printing is
concerned with the efficiency of
the visible word. So, too is the
practice of typographical design.
During the past century both
researchers and designers have
put forward proposals for
making printed letters
communicate more efficiently.
This report describes and
illustrates some of the more
significant of these proposals.
Seriffed typefaces with oblique axis of contrast tend to be more legible
Minion (top), Didot (bottom)
Handgloves
Handgloves
Non italicised words stand out within an italic text
Glynn, Britton & Tillman 1985
Legibility research in printing is concerned
with the efficiency of the visible word. So,
too is the practice of typographical design.
During the past century both researchers
and designers have put forward proposals
for making printed letters communicate
more efficiently. This report describes and
illustrates some of the more significant of
these proposals.
Weight changed noticed more than typeface change
Spencer, Reynolds & Coe 1975
Legibility research in printing is
concerned with the efficiency of the
visible word. So, too is the practice
of typographical design. During the
past century both researchers and
designers have put forward
proposals for making printed letters
communicate more efficiently. This
report describes and illustrates some
of the more significant of these
proposals.
Legibility research in printing is
concerned with the efficiency of the
visible word. So, too is the practice
of typographical design. During the
past century both researchers and
designers have put forward
proposals for making printed letters
communicate more efficiently. This
report describes and illustrates some
of the more significant of these
proposals.
Ambiguous letters
Burt 1959
C G
easily mistaken for
J f
h bGaramond italic
Baskerville
Imprint
Caslon italic
Plantin
easily mistaken for
easily mistaken for
Ambiguous letters
Ovink 1938
a
both acceptable g
t
amore legible than
g
tmore legible than
Ambiguous letters
Illinois
Illinois
Illinois
Ambiguous letters
III Ill 111
III Ill 111
III Ill 111
Gill Sans
Baskerville
ITC Officina Sans
Ambiguous letters
Illinois 111 millilitre Richard III
Illinois 111 millilitre Richard III
Illinois 111 millilitre Richard III
Illinois 111 millilitre Richard III
Illinois 111 millilitre Richard III
Illinois 111 millilitre Richard III
Gill Sans
Baskerville
Helvetica
Transport
ITC Officina Sans
Verdana
Medium or bold versions better in reverse
Wheildon 1995
Legibility research in printing is
concerned with the efficiency of
the visible word. So, too is the
practice of typographical design.
During the past century both
researchers and designers have
put forward proposals for making
printed letters communicate more
efficiently. This report describes
and illustrates some of the more
significant of these proposals.
Legibility research in printing is
concerned with the efficiency of
the visible word. So, too is the
practice of typographical design.
During the past century both
researchers and designers have
put forward proposals for making
printed letters communicate more
efficiently. This report describes
and illustrates some of the more
significant of these proposals.
Reversed continuous text may reduce
reading speed up to 15%
Holmes 1931,Taylor 1934
Readers may dislike it
Paterson & Tinker 1932
Measures: wide measures (long line lengths) cause regression
Continuous prose in Chinese is customarily set justified left and right. This is true even in classical poetry, where phrases have
set number of characters. The fact that Chinese characters are monowidth makes flush left, ragged right alignment impracti-
cal. Flush left alignment can only be achieved satisfactorily when line breaks are manually forced.
As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a sur-
face frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under
most circumstances. Column widths (hence line lengths) are calculated by number of characters.
Since Chinese is monowidth, and that column widths are measured by fixed numbers of characters, spacing should be
an easy matter. But this is not the case. There are three factors that make justification a complex issue: line-beginning and
line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of
spacing around certain punctuation marks.
Measure: wider leading helps, but not much if the type is too small
Continuous prose in Chinese is customarily set justified left and right. This is true even in classical poetry, where phrases have
set number of characters. The fact that Chinese characters are monowidth makes flush left, ragged right alignment impracti-
cal. Flush left alignment can only be achieved satisfactorily when line breaks are manually forced.
As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a sur-
face frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under
most circumstances. Column widths (hence line lengths) are calculated by number of characters.
Since Chinese is monowidth, and that column widths are measured by fixed numbers of characters, spacing should be
an easy matter. But this is not the case. There are three factors that make justification a complex issue: line-beginning and
line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of
spacing around certain punctuation marks.
Measure: still too wide
Continuous prose in Chinese is customarily set justified left and right. This is true even in classical
poetry, where phrases have set number of characters. The fact that Chinese characters are monow-
idth makes flush left, ragged right alignment impractical. Flush left alignment can only be achieved
satisfactorily when line breaks are manually forced.
As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character,
by way of defining a surface frame that is a smaller percentage than the body frame (em square).
This internal spacing is the optimal spacing under most circumstances. Column widths (hence line
lengths) are calculated by number of characters.
Since Chinese is monowidth, and that column widths are measured by fixed numbers of char-
acters, spacing should be an easy matter. But this is not the case. There are three factors that make
justification a complex issue: line-beginning and line-ending rules; the insertion of proportional
width characters, such as Latin glyphs, symbols or figures; adjustments of spacing around certain
punctuation marks.
Measure: just right (60–70 characters)
Continuous prose in Chinese is customarily set justified left and right.
This is true even in classical poetry, where phrases have set number of
characters. The fact that Chinese characters are monowidth makes flush
left, ragged right alignment impractical. Flush left alignment can only be
achieved satisfactorily when line breaks are manually forced.
As mentioned earlier, spacing in Chinese typefaces is achieved inter-
nally within each character, by way of defining a surface frame that is a
smaller percentage than the body frame (em square). This internal spacing
is the optimal spacing under most circumstances. Column widths (hence
line lengths) are calculated by number of characters.
Since Chinese is monowidth, and that column widths are measured
by fixed numbers of characters, spacing should be an easy matter. But this
is not the case. There are three factors that make justification a complex
issue: line-beginning and line-ending rules; the insertion of proportional
width characters, such as Latin glyphs, symbols or figures; adjustments of
spacing around certain punctuation marks.
Measure: still fine but a bit narrow
Continuous prose in Chinese is customarily set
justified left and right. This is true even in clas-
sical poetry, where phrases have set number of
characters. The fact that Chinese characters are
monowidth makes flush left, ragged right align-
ment impractical. Flush left alignment can only
be achieved satisfactorily when line breaks are
manually forced.
As mentioned earlier, spacing in Chinese
typefaces is achieved internally within each
character, by way of defining a surface frame
that is a smaller percentage than the body frame
(em square). This internal spacing is the optimal
spacing under most circumstances. Column
widths (hence line lengths) are calculated by
number of characters.
Since Chinese is monowidth, and that col-
umn widths are measured by fixed numbers of
characters, spacing should be an easy matter.
But this is not the case. There are three factors
that make justification a complex issue: line-be-
ginning and line-ending rules; the insertion of
proportional width characters, such as Latin
glyphs, symbols or figures; adjustments of spac-
ing around certain punctuation marks.
Measure: too narrow
Continuous prose in
Chinese is customarily
set justified left and
right. This is true even
in classical poetry,
where phrases have
set number of char-
acters. The fact that
Chinese characters
are monowidth makes
flush left, ragged right
alignment impractical.
Flush left alignment
can only be achieved
satisfactorily when
line breaks are manu-
ally forced.
As mentioned
earlier, spacing in
Chinese typefaces is
achieved internally
within each character,
by way of defining a
surface frame that is
a smaller percentage
Paragraph division: pilcrows
Continuous prose in Chinese is customarily set justified left and right. This
is true even in classical poetry, where phrases have set number of charac-
ters. The fact that Chinese characters are monowidth makes flush left, rag-
ged right alignment impractical. Flush left alignment can only be achieved
satisfactorily when line breaks are manually forced. ¶ As mentioned earlier,
spacing in Chinese typefaces is achieved internally within each character,
by way of defining a surface frame that is a smaller percentage than the body
frame (em square). This internal spacing is the optimal spacing under most
circumstances. Column widths (hence line lengths) are calculated by num-
ber of characters. ¶ Since Chinese is monowidth, and that column widths
are measured by fixed numbers of characters, spacing should be an easy
matter. But this is not the case. There are three factors that make justifica-
tion a complex issue: line-beginning and line-ending rules; the insertion
of proportional width characters, such as Latin glyphs, symbols or figures;
adjustments of spacing around certain punctuation marks.
Paragraph division: pilcrows
Continuous prose in Chinese is customarily set justified left and right. This
is true even in classical poetry, where phrases have set number of charac-
ters. The fact that Chinese characters are monowidth makes flush left, rag-
ged right alignment impractical. Flush left alignment can only be achieved
satisfactorily when line breaks are manually forced.
¶ As mentioned earlier, spacing in Chinese typefaces is achieved internally
within each character, by way of defining a surface frame that is a smaller
percentage than the body frame (em square). This internal spacing is the
optimal spacing under most circumstances. Column widths (hence line
lengths) are calculated by number of characters.
¶ Since Chinese is monowidth, and that column widths are measured by
fixed numbers of characters, spacing should be an easy matter. But this is
not the case. There are three factors that make justification a complex issue:
line-beginning and line-ending rules; the insertion of proportional width
characters, such as Latin glyphs, symbols or figures; adjustments of spacing
around certain punctuation marks.
Paragraph division: first-line indent (using leading value)
Continuous prose in Chinese is customarily set justified left and right. This
is true even in classical poetry, where phrases have set number of charac-
ters. The fact that Chinese characters are monowidth makes flush left, rag-
ged right alignment impractical. Flush left alignment can only be achieved
satisfactorily when line breaks are manually forced.
As mentioned earlier, spacing in Chinese typefaces is achieved internally
within each character, by way of defining a surface frame that is a smaller
percentage than the body frame (em square). This internal spacing is the
optimal spacing under most circumstances. Column widths (hence line
lengths) are calculated by number of characters.
Since Chinese is monowidth, and that column widths are measured by
fixed numbers of characters, spacing should be an easy matter. But this is
not the case. There are three factors that make justification a complex issue:
line-beginning and line-ending rules; the insertion of proportional width
characters, such as Latin glyphs, symbols or figures; adjustments of spacing
around certain punctuation marks.
Paragraph division: line breaks only
Continuous prose in Chinese is customarily set justified left and right. This
is true even in classical poetry, where phrases have set number of characters.
The fact that Chinese characters are monowidth makes flush left, ragged
right alignment impractical. Flush left alignment can only be achieved satis-
factorily when line breaks are manually forced.
As mentioned earlier, spacing in Chinese typefaces is achieved internally
within each character, by way of defining a surface frame that is a smaller
percentage than the body frame (em square). This internal spacing is the
optimal spacing under most circumstances. Column widths (hence line
lengths) are calculated by number of characters.
Since Chinese is monowidth, and that column widths are measured by fixed
numbers of characters, spacing should be an easy matter. But this is not
the case. There are three factors that make justification a complex issue:
line-beginning and line-ending rules; the insertion of proportional width
characters, such as Latin glyphs, symbols or figures; adjustments of spacing
around certain punctuation marks.
Paragraph division: line breaks only (flush-left, ragged right)
Continuous prose in Chinese is customarily set justified left and right. This
is true even in classical poetry, where phrases have set number of charac-
ters. The fact that Chinese characters are monowidth makes flush left, rag-
ged right alignment impractical. Flush left alignment can only be achieved
satisfactorily when line breaks are manually forced.
As mentioned earlier, spacing in Chinese typefaces is achieved internally
within each character, by way of defining a surface frame that is a smaller
percentage than the body frame (em square). This internal spacing is the
optimal spacing under most circumstances. Column widths (hence line
lengths) are calculated by number of characters.
Since Chinese is monowidth, and that column widths are measured by
fixed numbers of characters, spacing should be an easy matter. But this is
not the case. There are three factors that make justification a complex is-
sue: line-beginning and line-ending rules; the insertion of proportional
width characters, such as Latin glyphs, symbols or figures; adjustments of
spacing around certain punctuation marks.
Paragraph division: hanging indent (or called outdent)
Continuous prose in Chinese is customarily set justified left and right. This is
true even in classical poetry, where phrases have set number of characters.
The fact that Chinese characters are monowidth makes flush left, ragged
right alignment impractical. Flush left alignment can only be achieved satis-
factorily when line breaks are manually forced.
Asmentionedearlier,spacinginChinesetypefacesisachievedinternallywithin
each character, by way of defining a surface frame that is a smaller percent-
age than the body frame (em square). This internal spacing is the optimal
spacing under most circumstances. Column widths (hence line lengths) are
calculated by number of characters.
Since Chinese is monowidth, and that column widths are measured by fixed
numbers of characters, spacing should be an easy matter. But this is not
the case. There are three factors that make justification a complex issue:
line-beginning and line-ending rules; the insertion of proportional width
characters, such as Latin glyphs, symbols or figures; adjustments of spacing
around certain punctuation marks.
Paragraph division: line space (full line)
Continuous prose in Chinese is customarily set justified left and right. This
is true even in classical poetry, where phrases have set number of characters.
The fact that Chinese characters are monowidth makes flush left, ragged
right alignment impractical. Flush left alignment can only be achieved satis-
factorily when line breaks are manually forced.
As mentioned earlier, spacing in Chinese typefaces is achieved internally
within each character, by way of defining a surface frame that is a smaller
percentage than the body frame (em square). This internal spacing is the
optimal spacing under most circumstances. Column widths (hence line
lengths) are calculated by number of characters.
Since Chinese is monowidth, and that column widths are measured by fixed
numbers of characters, spacing should be an easy matter. But this is not
the case. There are three factors that make justification a complex issue:
line-beginning and line-ending rules; the insertion of proportional width
characters, such as Latin glyphs, symbols or figures; adjustments of spacing
around certain punctuation marks.
Paragraph division: line space (half a line)
Continuous prose in Chinese is customarily set justified left and right. This
is true even in classical poetry, where phrases have set number of characters.
The fact that Chinese characters are monowidth makes flush left, ragged
right alignment impractical. Flush left alignment can only be achieved satis-
factorily when line breaks are manually forced.
As mentioned earlier, spacing in Chinese typefaces is achieved internally
within each character, by way of defining a surface frame that is a smaller
percentage than the body frame (em square). This internal spacing is the
optimal spacing under most circumstances. Column widths (hence line
lengths) are calculated by number of characters.
Since Chinese is monowidth, and that column widths are measured by fixed
numbers of characters, spacing should be an easy matter. But this is not
the case. There are three factors that make justification a complex issue:
line-beginning and line-ending rules; the insertion of proportional width
characters, such as Latin glyphs, symbols or figures; adjustments of spacing
around certain punctuation marks.
Paragraph division: line space and first-line indent (redundant)
Continuous prose in Chinese is customarily set justified left and right. This
is true even in classical poetry, where phrases have set number of characters.
The fact that Chinese characters are monowidth makes flush left, ragged
right alignment impractical. Flush left alignment can only be achieved satis-
factorily when line breaks are manually forced.
As mentioned earlier, spacing in Chinese typefaces is achieved internally
within each character, by way of defining a surface frame that is a smaller
percentage than the body frame (em square). This internal spacing is the
optimal spacing under most circumstances. Column widths (hence line
lengths) are calculated by number of characters.
Since Chinese is monowidth, and that column widths are measured by
fixed numbers of characters, spacing should be an easy matter. But this is
not the case. There are three factors that make justification a complex issue:
line-beginning and line-ending rules; the insertion of proportional width
characters, such as Latin glyphs, symbols or figures; adjustments of spacing
around certain punctuation marks.

More Related Content

PPT
Fore Grounding & Interpretation By Nazma
PDF
Medieval period, communication and texts
PPT
Foregrounding2
PPTX
English report prose and poetry
PPT
Cohesion, Coherence and Textuality
PPT
4 figurative language
PDF
Type Compendium
PPT
Coherence And Cohesion
Fore Grounding & Interpretation By Nazma
Medieval period, communication and texts
Foregrounding2
English report prose and poetry
Cohesion, Coherence and Textuality
4 figurative language
Type Compendium
Coherence And Cohesion

What's hot (20)

DOCX
PPTX
The unit of translation and discourse analysis
PPT
Cohesion and coherence
PPT
Cohesion
PPTX
PPSX
Week4d pptslides writing with coherence
DOC
Textual cohesion
PPT
Cohesion And Coherence
PDF
Coherence and cohesion
PPTX
Cohesion and coherence
PDF
Lecture 75 Interpreting Discourse
PPTX
Coherence
DOC
A Pragmatic Analysis of Verbal Irony in Pride and Prejudice
PPT
Cohesion Types
PPTX
The nature of reference in text and in powerpoint presentation
DOC
Cohesion In English
DOC
Kris' Dissertation Chapter 4 Methodology and Data Collection
PPTX
Cohesion presentation
ODP
Presentation of text linguistics
PPTX
Stylistic analysis
The unit of translation and discourse analysis
Cohesion and coherence
Cohesion
Week4d pptslides writing with coherence
Textual cohesion
Cohesion And Coherence
Coherence and cohesion
Cohesion and coherence
Lecture 75 Interpreting Discourse
Coherence
A Pragmatic Analysis of Verbal Irony in Pride and Prejudice
Cohesion Types
The nature of reference in text and in powerpoint presentation
Cohesion In English
Kris' Dissertation Chapter 4 Methodology and Data Collection
Cohesion presentation
Presentation of text linguistics
Stylistic analysis
Ad

Similar to Typographic hierarchy and legibility research (20)

PDF
pdfcoffee.com_a-guide-to-korean-characters-hanja-pdf-free.pdf
PPTX
Demystifying Mandarin - Learn Chinese by Hutong School
PPT
Writing history progretion
PDF
Typographic Design
PPTX
semantic field in fairy tales
PDF
Shakespeare for English Language Learners
PPTX
Ll1 structure
PPTX
History of Translation 1980's era
PPTX
Presentasi Literatur Bahasa Inggris sangat penting bagi mereka yang ingin mem...
PPTX
Some Issues of Contention.
PPTX
NLP_KASHK:Context-Free Grammar for English
PDF
A Contrastive Analysis Between the Chinese and English Versions of If I Were...
PPT
Stylistics 3 The Purpose of Stylistics.ppt
PPTX
FIGURES-OF-SPEECH.pptx this will help those students who did not know the fig...
PPTX
Intro. to Stylistics
PPTX
Special problems in literary translation
PDF
The Apatani Alphabet
DOCX
Of myths, overarching generalizations, and grains of truth
PPTX
Ppt essay cultural heritage
pdfcoffee.com_a-guide-to-korean-characters-hanja-pdf-free.pdf
Demystifying Mandarin - Learn Chinese by Hutong School
Writing history progretion
Typographic Design
semantic field in fairy tales
Shakespeare for English Language Learners
Ll1 structure
History of Translation 1980's era
Presentasi Literatur Bahasa Inggris sangat penting bagi mereka yang ingin mem...
Some Issues of Contention.
NLP_KASHK:Context-Free Grammar for English
A Contrastive Analysis Between the Chinese and English Versions of If I Were...
Stylistics 3 The Purpose of Stylistics.ppt
FIGURES-OF-SPEECH.pptx this will help those students who did not know the fig...
Intro. to Stylistics
Special problems in literary translation
The Apatani Alphabet
Of myths, overarching generalizations, and grains of truth
Ppt essay cultural heritage
Ad

Recently uploaded (20)

PDF
Pongal 2026 Sponsorship Presentation - Bhopal Tamil Sangam
PPTX
CLASS_11_BUSINESS_STUDIES_PPT_CHAPTER_1_Business_Trade_Commerce.pptx
PDF
Architecture Design Portfolio- VICTOR OKUTU
PPTX
Entrepreneur intro, origin, process, method
PPTX
a group casestudy on architectural aesthetic and beauty
PDF
321 LIBRARY DESIGN.pdf43354445t6556t5656
PPT
EthicsNotesSTUDENTCOPYfghhnmncssssx sjsjsj
PPTX
22CDO02-IMGD-UNIT-I-MOBILE GAME DESIGN PROCESS
PPTX
Acoustics new for. Sound insulation and absorber
PPTX
EDP Competencies-types, process, explanation
PPTX
Orthtotics presentation regarding physcial therapy
PDF
Design Thinking - Module 1 - Introduction To Design Thinking - Dr. Rohan Dasg...
PDF
Skskkxiixijsjsnwkwkaksixindndndjdjdjsjjssk
PDF
The Basics of Presentation Design eBook by VerdanaBold
PPT
aksharma-dfs.pptgfgfgdfgdgdfgdfgdgdrgdgdgdgdgdgadgdgd
PPTX
DOC-20250430-WA0014._20250714_235747_0000.pptx
PDF
Interior Structure and Construction A1 NGYANQI
PPTX
NEW EIA PART B - Group 5 (Section 50).pptx
PPTX
Evolution_of_Computing_Presentation (1).pptx
PPTX
rapid fire quiz in your house is your india.pptx
Pongal 2026 Sponsorship Presentation - Bhopal Tamil Sangam
CLASS_11_BUSINESS_STUDIES_PPT_CHAPTER_1_Business_Trade_Commerce.pptx
Architecture Design Portfolio- VICTOR OKUTU
Entrepreneur intro, origin, process, method
a group casestudy on architectural aesthetic and beauty
321 LIBRARY DESIGN.pdf43354445t6556t5656
EthicsNotesSTUDENTCOPYfghhnmncssssx sjsjsj
22CDO02-IMGD-UNIT-I-MOBILE GAME DESIGN PROCESS
Acoustics new for. Sound insulation and absorber
EDP Competencies-types, process, explanation
Orthtotics presentation regarding physcial therapy
Design Thinking - Module 1 - Introduction To Design Thinking - Dr. Rohan Dasg...
Skskkxiixijsjsnwkwkaksixindndndjdjdjsjjssk
The Basics of Presentation Design eBook by VerdanaBold
aksharma-dfs.pptgfgfgdfgdgdfgdfgdgdrgdgdgdgdgdgadgdgd
DOC-20250430-WA0014._20250714_235747_0000.pptx
Interior Structure and Construction A1 NGYANQI
NEW EIA PART B - Group 5 (Section 50).pptx
Evolution_of_Computing_Presentation (1).pptx
rapid fire quiz in your house is your india.pptx

Typographic hierarchy and legibility research

  • 1. Twyman, M (1981)
 ‘Typography without words’,
 in Visible Language, XV no. 1, 5–12 Spencer, Herbert (1969)
 The visible word
 London: Lund Humphries/Royal College of Art Typographic hierarchy Legibility research
  • 2. Twyman 1981 A notation system to guide discussion and design decisions Notation that describes function rather than exact form (making use of the technical limitations of a typewriter) Spatial grouping (Gestalt psychology): vertical and lateral spatial cues To show the typography is a modular system About the relationship between levels of heading and paragraphs of continuous text
  • 3. Which examples would be ‘reasonably unambiguous in visual terms’? A xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx ooooo xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx ooooo xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx D xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx ooooo xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx ooooo xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx B xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx ooooo xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx ooooo xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx E xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx ooooo xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx ooooo xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx C xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx ooooo xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx oooo xxxxxxxxxxxxxxxxxxxxx ooo xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx F xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx oooo xxxxxxxxxxxxxxxxxxxxx ooo xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx ooooo xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx
  • 4. Which examples would be ‘reasonably unambiguous in visual terms’? largest ethnic group in East Asia. Other languages also use variant forms of Hanzi: Kanji (Japanese), Hanja (Korean), and Chữ Nôm (Vietnamese). The Chinese language There are eight main varieties of speech in China thataredistinctlydifferentfromeachotherinterms of pronunciation and vocabulary, though they are often referred to as ‘dialects’ rather than languages. Linguistic aspects of Hanzi Chinese characters are classified as logographic, meaning that each character represents a mor- pheme, the smallest unit of meaning. It is often said that Chinese characters are pictographic; this is only partly true. largest ethnic group in East Asia. Other languages also use variant forms of Hanzi: Kanji (Japanese), Hanja (Korean), and Chữ Nôm (Vietnamese). The Chinese language There are eight main varieties of speech in China thataredistinctlydifferentfromeachotherinterms of pronunciation and vocabulary, though they are often referred to as ‘dialects’ rather than languages. Linguistic aspects of Hanzi Chinese characters are classified as logographic, meaning that each character represents a mor- pheme, the smallest unit of meaning. It is often said that Chinese characters are pictographic; this is only partly true. largest ethnic group in East Asia. Other languages also use variant forms of Hanzi: Kanji (Japanese), Hanja (Korean), and Chữ Nôm (Vietnamese). The Chinese language There are eight main varieties of speech in China thataredistinctlydifferentfromeachotherinterms of pronunciation and vocabulary, though they are often referred to as ‘dialects’ rather than languages. Linguistic aspects of Hanzi Chinese characters are classified as logographic, meaning that each character represents a mor- pheme, the smallest unit of meaning. It is often said that Chinese characters are pictographic; this is only partly true. largest ethnic group in East Asia. Other languages also use variant forms of Hanzi: Kanji (Japanese), Hanja (Korean), and Chữ Nôm (Vietnamese). The Chinese language There are eight main varieties of speech in China thataredistinctlydifferentfromeachotherinterms of pronunciation and vocabulary, though they are often referred to as ‘dialects’ rather than languages. Linguistic aspects of Hanzi Chinese characters are classified as logographic, meaning that each character represents a mor- pheme, the smallest unit of meaning. It is often said that Chinese characters are pictographic; this is only partly true. A B D E ✔ ✔
  • 5. Which examples would be ‘reasonably unambiguous in visual terms’? largest ethnic group in East Asia. Other languages also use variant forms of Hanzi: Kanji (Japanese), Hanja (Korean), and Chữ Nôm (Vietnamese). The Chinese language There are eight main varieties of speech in China thataredistinctlydifferentfromeachotherinterms of pronunciation and vocabulary, though they are often referred to as ‘dialects’ rather than languages. Chinese characters are classified as logographic, meaning that each character represents a mor- pheme, the smallest unit of meaning. It is often said that Chinese characters are pictographic; this is only partly true. Linguistic aspects of Hanzi largest ethnic group in East Asia. Other languages also use variant forms of Hanzi: Kanji (Japanese), Hanja (Korean), and Chữ Nôm (Vietnamese). There are eight main varieties of speech in China thataredistinctlydifferentfromeachotherinterms of pronunciation and vocabulary, though they are often referred to as ‘dialects’ rather than languages. Linguistic aspects of Hanzi Chinese characters are classified as logographic, meaning that each character represents a mor- pheme, the smallest unit of meaning. It is often said that Chinese characters are pictographic; this is only partly true. The Chinese language C F
  • 6. Which example articulate the text as clearly as possible? A xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx ooooo xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx B xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx ooooo xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx C xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx ooooo xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx D xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx ooooo xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx
  • 7. Which example articulate the text as clearly as possible? Continuous prose in Chinese is customarily set justified left and right. This is true even in clas- sical poetry, where phrases have set number of characters. The fact that Chinese characters are monowidth makes flush left, ragged right align- ment impractical. Spacing and layout As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a surface frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under most circumstances.Columnwidths(hencelinelengths) are calculated by number of characters. Since Chinese is monowidth, and that column widths are measured by fixed numbers of charac- ters, spacing should be an easy matter. But this is not the case. There are three factors that make justification a complex issue: line-beginning and line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of spacing around certain punctuation marks. Continuous prose in Chinese is customarily set justified left and right. This is true even in clas- sical poetry, where phrases have set number of characters. The fact that Chinese characters are monowidth makes flush left, ragged right align- ment impractical. Spacing and layout As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a surface frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under most circumstances.Columnwidths(hencelinelengths) are calculated by number of characters. Since Chinese is monowidth, and that column widths are measured by fixed numbers of char- acters, spacing should be an easy matter. But this is not the case. There are three factors that make justification a complex issue: line-beginning and line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of spacing around certain punctuation marks. A B
  • 8. Which example articulate the text as clearly as possible? Continuous prose in Chinese is customarily set justified left and right. This is true even in clas- sical poetry, where phrases have set number of characters. The fact that Chinese characters are monowidth makes flush left, ragged right align- ment impractical. Spacing and layout As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a surface frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under most circumstances.Columnwidths(hencelinelengths) are calculated by number of characters. Since Chinese is monowidth, and that column widths are measured by fixed numbers of charac- ters, spacing should be an easy matter. But this is not the case. There are three factors that make justification a complex issue: line-beginning and line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of spacing around certain punctuation marks. Continuous prose in Chinese is customarily set justified left and right. This is true even in clas- sical poetry, where phrases have set number of characters. The fact that Chinese characters are monowidth makes flush left, ragged right align- ment impractical. Spacing and layout As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a surface frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under most circumstances.Columnwidths(hencelinelengths) are calculated by number of characters. Since Chinese is monowidth, and that column widths are measured by fixed numbers of charac- ters, spacing should be an easy matter. But this is not the case. There are three factors that make justification a complex issue: line-beginning and line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of spacing around certain C D ✔
  • 9. moulds to mass produce individual types, which is an essential criteria for the development of typefaces. Chinese (Hanzi) Chinese is currently the most widely used language in the world, used by 1302 million people (Lewis et al 2016). The Chinese language is written in Hanzi, where Han refers to the dynasty (202 BC–220 AD) as well as the largest ethnic group in East Asia. Other languages also use variant forms of Hanzi: Kanji (Japanese), Hanja (Korean), and Chữ Nôm (Vietnamese). The Chinese language and the Hanzi script There are eight main varieties of speech in China that are distinctly dif- ferent from each other in terms of pronunciation and vocabulary, though they are often referred to as ‘dialects’ rather than languages. Linguistic aspects of Hanzi Chinese characters are classified as logographic, meaning that each char- acter represents a morpheme, the smallest unit of meaning. It is often said that Chinese characters are pictographic; this is only partly true. Traditional and Simplified Chinese There are currently two forms of Chinese orthography: Traditional Chinese and Simplified Chinese. Simplified Chinese characters were derived from the traditional forms by reducing the number of strokes, and were first proposed during the Republic of China era in 1935. Anatomical characteristics of Chinese characters Chinese characters are customarily monowidth; they sit within a notional square. In typesetting, this notional square is defined as the em square, How many orders of heading? xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx ooooo xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx ooooo xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx iiii xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx iiii xxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxx ooooo xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx 1 2 3 4
  • 10. is therefore important that justification is done via Chinese justification rules rather than Latin ones to ensure good spacing. Code-mixing Code-mixing is a term borrowed from linguistics. It is referred to the in- sertion of words and short phrases from another script into the text flow of the main script. Explanation or disambiguation Chinese translations of technical or specialised terms may be ambiguous because they are little known or non-standard. This may also be true for names of places and foreign people. Parallel typesetting Text extent A passage translated from English into Chinese is usually shorter. This is due to the fact that Chinese is a more concise writing system. Typesize The anatomical differences between Chinese and Latin characters make them appear different in size even when they are both set in the same point size. Chinese characters appear larger, as they occupy the em square more fully than Latin characters. Weight and density Chinese characters can have anywhere between one to more than 30 strokes, making their densities vary quite considerably. A page of Chinese text can therefore appear rather spotty Which is the second order of heading? xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx ooooo xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx iiii xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx ooooo ooooo xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx iiii xxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx iiii xxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxx
  • 11. Content structure The norm: the main text elements above the norm elements below the norm eg chapter headings
 subheads
 images
 introductory text
 pull quotes
 etc . . . eg footnotes
 captions
 margin notes
 page numbers
 headers and footers
 etc . . . continuous text
 body text
  • 12. Typographic variations for headings: differentiation from the norm Continuous text: roman Heading: italic Heading: small caps HEADING: ALL CAPS Heading: bold Heading: another typeface Heading: colour Heading: size
  • 13. Heading differentiation: line break and space above each glyph, with a slightly smaller square defining the margin within which the character sits, referred to as the ‘surface frame’ (Lu and Tang 2016:111). Typeface designers define surface frames as percentages of the body frame. The larger this percentage, the more tightly spaced the typeface appear. Units of measurement The Western point system was brought into China along with the in- dustrialised process of printing with moveable lead type in 1859–60 by Presbyterian missionary William Gamble. A modular size system with nu- merical labels were subsequently adopted for Chinese lead type, with nine related sizes which were mapped onto points. The most common size for continuous text was number 5, equal to 10.5 point. Latin characters were casted in this size for the purpose of code-mixed setting. Chinese digi- tal typesetting now adopts the Western system of points and picas. The concept of the em is important in Chinese typesetting, as column widths, spacing and indents are defined using fractions or multiples of the em, which is the point size of the current type used.
  • 14. Heading differentiation: equal space above and below (not ideal) each glyph, with a slightly smaller square defining the margin within which the character sits, referred to as the ‘surface frame’ (Lu and Tang 2016:111). Typeface designers define surface frames as percentages of the body frame. The larger this percentage, the more tightly spaced the typeface appear. Units of measurement The Western point system was brought into China along with the in- dustrialised process of printing with moveable lead type in 1859–60 by Presbyterian missionary William Gamble. A modular size system with nu- merical labels were subsequently adopted for Chinese lead type, with nine related sizes which were mapped onto points. The most common size for continuous text was number 5, equal to 10.5 point. Latin characters were casted in this size for the purpose of code-mixed setting. Chinese digi- tal typesetting now adopts the Western system of points and picas. The concept of the em is important in Chinese typesetting, as column widths, spacing and indents are defined using fractions or multiples of the em, which is the point size of the current type used.
  • 15. Heading differentiation: closer to paragraph below than above each glyph, with a slightly smaller square defining the margin within which the character sits, referred to as the ‘surface frame’ (Lu and Tang 2016:111). Typeface designers define surface frames as percentages of the body frame. The larger this percentage, the more tightly spaced the typeface appear. Units of measurement The Western point system was brought into China along with the in- dustrialised process of printing with moveable lead type in 1859–60 by Presbyterian missionary William Gamble. A modular size system with nu- merical labels were subsequently adopted for Chinese lead type, with nine related sizes which were mapped onto points. The most common size for continuous text was number 5, equal to 10.5 point. Latin characters were casted in this size for the purpose of code-mixed setting. Chinese digi- tal typesetting now adopts the Western system of points and picas. The concept of the em is important in Chinese typesetting, as column widths, spacing and indents are defined using fractions or multiples of the em, which is the point size of the current type used.
  • 16. Heading differentiation: now bold each glyph, with a slightly smaller square defining the margin within which the character sits, referred to as the ‘surface frame’ (Lu and Tang 2016:111). Typeface designers define surface frames as percentages of the body frame. The larger this percentage, the more tightly spaced the typeface appear. Units of measurement The Western point system was brought into China along with the in- dustrialised process of printing with moveable lead type in 1859–60 by Presbyterian missionary William Gamble. A modular size system with nu- merical labels were subsequently adopted for Chinese lead type, with nine related sizes which were mapped onto points. The most common size for continuous text was number 5, equal to 10.5 point. Latin characters were casted in this size for the purpose of code-mixed setting. Chinese digi- tal typesetting now adopts the Western system of points and picas. The concept of the em is important in Chinese typesetting, as column widths, spacing and indents are defined using fractions or multiples of the em, which is the point size of the current type used.
  • 17. Heading differentiation: bold and left indent each glyph, with a slightly smaller square defining the margin within which the character sits, referred to as the ‘surface frame’ (Lu and Tang 2016:111). Typeface designers define surface frames as percentages of the body frame. The larger this percentage, the more tightly spaced the typeface appear. Units of measurement The Western point system was brought into China along with the in- dustrialised process of printing with moveable lead type in 1859–60 by Presbyterian missionary William Gamble. A modular size system with nu- merical labels were subsequently adopted for Chinese lead type, with nine related sizes which were mapped onto points. The most common size for continuous text was number 5, equal to 10.5 point. Latin characters were casted in this size for the purpose of code-mixed setting. Chinese digi- tal typesetting now adopts the Western system of points and picas. The concept of the em is important in Chinese typesetting, as column widths, spacing and indents are defined using fractions or multiples of the em, which is the point size of the current type used.
  • 18. Heading differentiation: larger and left indent each glyph, with a slightly smaller square defining the margin within which the character sits, referred to as the ‘surface frame’ (Lu and Tang 2016:111). Typeface designers define surface frames as percentages of the body frame. The larger this percentage, the more tightly spaced the typeface appear. Units of measurement The Western point system was brought into China along with the in- dustrialised process of printing with moveable lead type in 1859–60 by Presbyterian missionary William Gamble. A modular size system with nu- merical labels were subsequently adopted for Chinese lead type, with nine related sizes which were mapped onto points. The most common size for continuous text was number 5, equal to 10.5 point. Latin characters were casted in this size for the purpose of code-mixed setting. Chinese digi- tal typesetting now adopts the Western system of points and picas. The concept of the em is important in Chinese typesetting, as column widths, spacing and indents are defined using fractions or multiples of the em, which is the point size of the current type used.
  • 19. Heading differentiation: larger and left indent (smaller line spaces) each glyph, with a slightly smaller square defining the margin within which the character sits, referred to as the ‘surface frame’ (Lu and Tang 2016:111). Typeface designers define surface frames as percentages of the body frame. The larger this percentage, the more tightly spaced the typeface appear. Units of measurement The Western point system was brought into China along with the in- dustrialised process of printing with moveable lead type in 1859–60 by Presbyterian missionary William Gamble. A modular size system with nu- merical labels were subsequently adopted for Chinese lead type, with nine related sizes which were mapped onto points. The most common size for continuous text was number 5, equal to 10.5 point. Latin characters were casted in this size for the purpose of code-mixed setting. Chinese digi- tal typesetting now adopts the Western system of points and picas. The concept of the em is important in Chinese typesetting, as column widths, spacing and indents are defined using fractions or multiples of the em, which is the point size of the current type used.
  • 20. Differentiation of heading: larger and outdent each glyph, with a slightly smaller square defining the margin within which the character sits, referred to as the ‘surface frame’ (Lu and Tang 2016:111). Typeface designers define surface frames as percentages of the body frame. The larger this percentage, the more tightly spaced the typeface appear. Units of measurement The Western point system was brought into China along with the in- dustrialised process of printing with moveable lead type in 1859–60 by Presbyterian missionary William Gamble. A modular size system with nu- merical labels were subsequently adopted for Chinese lead type, with nine related sizes which were mapped onto points. The most common size for continuous text was number 5, equal to 10.5 point. Latin characters were casted in this size for the purpose of code-mixed setting. Chinese digi- tal typesetting now adopts the Western system of points and picas. The concept of the em is important in Chinese typesetting, as column widths, spacing and indents are defined using fractions or multiples of the em, which is the point size of the current type used.
  • 21. may occur when words run together. This should be avoided when writing and editing the text. Punctuation and typographic treatments can also be used to avoid ambiguity. Anatomical characteristics of Chinese characters Chinese characters are customarily monowidth; they sit within a notional square. In typesetting, this notional square is defined as the em square, referred to as the ‘body frame’. Spacing between Chinese characters are defined internally within each glyph, with a slightly smaller square defin- ing the margin within which the character sits, referred to as the ‘surface frame’ (Lu and Tang 2016:111). Typeface designers define surface frames as percentages of the body frame. The larger this percentage, the more tightly spaced the typeface appear. Units of measurement The Western point system was brought into China along with the in- dustrialised process of printing with moveable lead type in 1859–60 by Presbyterian missionary William Gamble. A modular size system with nu- merical labels were subsequently adopted for Chinese lead type, with nine related sizes which were mapped onto points. The most common size for [paragraph] [heading 1] [paragraph] [heading 2]
  • 22. Spencer 1969 Review of literature on legibility research of over 100 years Empirical studies conducted by researchers outside the typography discipline Controlled experiments in lab situations Studies that guided typographic design
  • 23. Spencer 1969 Distinctiveness of characters Legibility of typefaces Importance of x-height rather than point size Capitals, lowercase, bold, italics, numerals, punctuation Interrelationship of type size, measure and leading Indication of paragraphs and alignment (line endings) Congeniality, atmosphere value and typographic allusion
  • 24. Legibility research: methods Distance Speed of perception (tachistoscope) Eye-movement Blink-rate Visual fatigue Peripheral vision Focal variation Visibility Heart-rate measurement Haploscope (stereoscope) Shaking table Rate of work
  • 25. We read in series of jumps along a line Javal 1878 The quick brown fox jumps over the lazy dog. Five boxing gymnasts jump fixations (pauses): where perception occurs 94% of reading time devoted to pauses each pause ¼ second saccades (jumps)
  • 26. x-height is as important if not more important than size for legibility Poulton 1955 Handgloves HandglovesGill Sans: smaller x-height Helvetica: larger x-height
  • 27. x-height is as important if not more important than size for legibility Poulton 1955 Gill Sans 170pt Helvetica 150pt Handgloves Handgloves x-heights have been equalised
  • 29. Serif and sans serif typefaces equally preferred by readers and read Hartley & Room 1983, Tinker 1963, Gould et al 1987, Zachrisson 1965 Legibility research in printing is concerned with the efficiency of the visible word. So, too is the practice of typographical design. During the past century both researchers and designers have put forward proposals for making printed letters communicate more efficiently. This report describes and illustrates some of the more significant of these proposals. Legibility research in printing is concerned with the efficiency of the visible word. So, too is the practice of typographical design. During the past century both researchers and designers have put forward proposals for making printed letters communicate more efficiently. This report describes and illustrates some of the more significant of these proposals. Serif faces may be easier to read in continuous text than sans serif faces Burt 1959, Hvistendal & Kahl 1975, Robinson, Abbamonte & Evans 1971,Wheeldon 1995
  • 30. All capitals setting slows down reading speed by 13–20% Breland & Breland 1944 Legibility research in printing is concerned with the efficiency of the visible word. So, too is the practice of typographical design. During the past century both researchers and designers have put forward proposals for making printed letters communicate more efficiently. This report describes and illustrates some of the more significant of these proposals. LEGIBILITY RESEARCH IN PRINTING IS CONCERNED WITH THE EFFICIENCY OF THE VISIBLE WORD. SO, TOO IS THE PRACTICE OF TYPOGRAPHICAL DESIGN. DURING THE PAST CENTURY BOTH RESEARCHERS AND DESIGNERS HAVE PUT FORWARD PROPOSALS FOR MAKING PRINTED LETTERS COMMUNICATE MORE EFFICIENTLY. THIS REPORT DESCRIBES AND ILLUSTRATES SOME OF THE MORE SIGNIFICANT OF THESE PROPOSALS.
  • 31. Word shapes are more distinctive in upper- and lowercase setting Ireland 1944; Paterson and Tinker 1928, 1940; Starch 1914 Handgloves HANDGLOVES
  • 32. Top halves of lowercase letters easier to recognise than lower halves Messmer 1903 Handgloves Handgloves
  • 33. Italic in continuous text can reduce reading speed substantially Foster & Bruce 1982, Tinker 1955 Legibility research in printing is concerned with the efficiency of the visible word. So, too is the practice of typographical design. During the past century both researchers and designers have put forward proposals for making printed letters communicate more efficiently. This report describes and illustrates some of the more significant of these proposals. Legibility research in printing is concerned with the efficiency of the visible word. So, too is the practice of typographical design. During the past century both researchers and designers have put forward proposals for making printed letters communicate more efficiently. This report describes and illustrates some of the more significant of these proposals.
  • 34. Readers preferred faces ‘approaching the appearance of bold face type’ Paterson and Tinker 1940 Legibility research in printing is concerned with the efficiency of the visible word. So, too is the practice of typographical design. During the past century both researchers and designers have put forward proposals for making printed letters communicate more efficiently. This report describes and illustrates some of the more significant of these proposals. Legibility research in printing is concerned with the efficiency of the visible word. So, too is the practice of typographical design. During the past century both researchers and designers have put forward proposals for making printed letters communicate more efficiently. This report describes and illustrates some of the more significant of these proposals.
  • 35. 1–4 points of leading read faster than set solid (no leading) Becker, Heinrich, van Sichowsny & Wendt 1970, Bentley 1921, Paterson & Tinker 1932 Legibility research in printing is concerned with the efficiency of the visible word. So, too is the practice of typographical design. During the past century both researchers and designers have put forward proposals for making printed letters communicate more efficiently. This report describes and illustrates some of the more significant of these proposals. Legibility research in printing is concerned with the efficiency of the visible word. So, too is the practice of typographical design. During the past century both researchers and designers have put forward proposals for making printed letters communicate more efficiently. This report describes and illustrates some of the more significant of these proposals.
  • 36. Lower contrast oblique axis (Minion) / high contrast vertical axis (Didot) Minion (left), Didot (right) / Wiggins 1967 (Anisson 1790s) Legibility research in printing is concerned with the efficiency of the visible word. So, too is the practice of typographical design. During the past century both researchers and designers have put forward proposals for making printed letters communicate more efficiently. This report describes and illustrates some of the more significant of these proposals. Legibility research in printing is concerned with the efficiency of the visible word. So, too is the practice of typographical design. During the past century both researchers and designers have put forward proposals for making printed letters communicate more efficiently. This report describes and illustrates some of the more significant of these proposals.
  • 37. Seriffed typefaces with oblique axis of contrast tend to be more legible Minion (top), Didot (bottom) Handgloves Handgloves
  • 38. Non italicised words stand out within an italic text Glynn, Britton & Tillman 1985 Legibility research in printing is concerned with the efficiency of the visible word. So, too is the practice of typographical design. During the past century both researchers and designers have put forward proposals for making printed letters communicate more efficiently. This report describes and illustrates some of the more significant of these proposals.
  • 39. Weight changed noticed more than typeface change Spencer, Reynolds & Coe 1975 Legibility research in printing is concerned with the efficiency of the visible word. So, too is the practice of typographical design. During the past century both researchers and designers have put forward proposals for making printed letters communicate more efficiently. This report describes and illustrates some of the more significant of these proposals. Legibility research in printing is concerned with the efficiency of the visible word. So, too is the practice of typographical design. During the past century both researchers and designers have put forward proposals for making printed letters communicate more efficiently. This report describes and illustrates some of the more significant of these proposals.
  • 40. Ambiguous letters Burt 1959 C G easily mistaken for J f h bGaramond italic Baskerville Imprint Caslon italic Plantin easily mistaken for easily mistaken for
  • 41. Ambiguous letters Ovink 1938 a both acceptable g t amore legible than g tmore legible than
  • 43. Ambiguous letters III Ill 111 III Ill 111 III Ill 111 Gill Sans Baskerville ITC Officina Sans
  • 44. Ambiguous letters Illinois 111 millilitre Richard III Illinois 111 millilitre Richard III Illinois 111 millilitre Richard III Illinois 111 millilitre Richard III Illinois 111 millilitre Richard III Illinois 111 millilitre Richard III Gill Sans Baskerville Helvetica Transport ITC Officina Sans Verdana
  • 45. Medium or bold versions better in reverse Wheildon 1995 Legibility research in printing is concerned with the efficiency of the visible word. So, too is the practice of typographical design. During the past century both researchers and designers have put forward proposals for making printed letters communicate more efficiently. This report describes and illustrates some of the more significant of these proposals. Legibility research in printing is concerned with the efficiency of the visible word. So, too is the practice of typographical design. During the past century both researchers and designers have put forward proposals for making printed letters communicate more efficiently. This report describes and illustrates some of the more significant of these proposals. Reversed continuous text may reduce reading speed up to 15% Holmes 1931,Taylor 1934 Readers may dislike it Paterson & Tinker 1932
  • 46. Measures: wide measures (long line lengths) cause regression Continuous prose in Chinese is customarily set justified left and right. This is true even in classical poetry, where phrases have set number of characters. The fact that Chinese characters are monowidth makes flush left, ragged right alignment impracti- cal. Flush left alignment can only be achieved satisfactorily when line breaks are manually forced. As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a sur- face frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under most circumstances. Column widths (hence line lengths) are calculated by number of characters. Since Chinese is monowidth, and that column widths are measured by fixed numbers of characters, spacing should be an easy matter. But this is not the case. There are three factors that make justification a complex issue: line-beginning and line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of spacing around certain punctuation marks.
  • 47. Measure: wider leading helps, but not much if the type is too small Continuous prose in Chinese is customarily set justified left and right. This is true even in classical poetry, where phrases have set number of characters. The fact that Chinese characters are monowidth makes flush left, ragged right alignment impracti- cal. Flush left alignment can only be achieved satisfactorily when line breaks are manually forced. As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a sur- face frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under most circumstances. Column widths (hence line lengths) are calculated by number of characters. Since Chinese is monowidth, and that column widths are measured by fixed numbers of characters, spacing should be an easy matter. But this is not the case. There are three factors that make justification a complex issue: line-beginning and line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of spacing around certain punctuation marks.
  • 48. Measure: still too wide Continuous prose in Chinese is customarily set justified left and right. This is true even in classical poetry, where phrases have set number of characters. The fact that Chinese characters are monow- idth makes flush left, ragged right alignment impractical. Flush left alignment can only be achieved satisfactorily when line breaks are manually forced. As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a surface frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under most circumstances. Column widths (hence line lengths) are calculated by number of characters. Since Chinese is monowidth, and that column widths are measured by fixed numbers of char- acters, spacing should be an easy matter. But this is not the case. There are three factors that make justification a complex issue: line-beginning and line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of spacing around certain punctuation marks.
  • 49. Measure: just right (60–70 characters) Continuous prose in Chinese is customarily set justified left and right. This is true even in classical poetry, where phrases have set number of characters. The fact that Chinese characters are monowidth makes flush left, ragged right alignment impractical. Flush left alignment can only be achieved satisfactorily when line breaks are manually forced. As mentioned earlier, spacing in Chinese typefaces is achieved inter- nally within each character, by way of defining a surface frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under most circumstances. Column widths (hence line lengths) are calculated by number of characters. Since Chinese is monowidth, and that column widths are measured by fixed numbers of characters, spacing should be an easy matter. But this is not the case. There are three factors that make justification a complex issue: line-beginning and line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of spacing around certain punctuation marks.
  • 50. Measure: still fine but a bit narrow Continuous prose in Chinese is customarily set justified left and right. This is true even in clas- sical poetry, where phrases have set number of characters. The fact that Chinese characters are monowidth makes flush left, ragged right align- ment impractical. Flush left alignment can only be achieved satisfactorily when line breaks are manually forced. As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a surface frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under most circumstances. Column widths (hence line lengths) are calculated by number of characters. Since Chinese is monowidth, and that col- umn widths are measured by fixed numbers of characters, spacing should be an easy matter. But this is not the case. There are three factors that make justification a complex issue: line-be- ginning and line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of spac- ing around certain punctuation marks.
  • 51. Measure: too narrow Continuous prose in Chinese is customarily set justified left and right. This is true even in classical poetry, where phrases have set number of char- acters. The fact that Chinese characters are monowidth makes flush left, ragged right alignment impractical. Flush left alignment can only be achieved satisfactorily when line breaks are manu- ally forced. As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a surface frame that is a smaller percentage
  • 52. Paragraph division: pilcrows Continuous prose in Chinese is customarily set justified left and right. This is true even in classical poetry, where phrases have set number of charac- ters. The fact that Chinese characters are monowidth makes flush left, rag- ged right alignment impractical. Flush left alignment can only be achieved satisfactorily when line breaks are manually forced. ¶ As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a surface frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under most circumstances. Column widths (hence line lengths) are calculated by num- ber of characters. ¶ Since Chinese is monowidth, and that column widths are measured by fixed numbers of characters, spacing should be an easy matter. But this is not the case. There are three factors that make justifica- tion a complex issue: line-beginning and line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of spacing around certain punctuation marks.
  • 53. Paragraph division: pilcrows Continuous prose in Chinese is customarily set justified left and right. This is true even in classical poetry, where phrases have set number of charac- ters. The fact that Chinese characters are monowidth makes flush left, rag- ged right alignment impractical. Flush left alignment can only be achieved satisfactorily when line breaks are manually forced. ¶ As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a surface frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under most circumstances. Column widths (hence line lengths) are calculated by number of characters. ¶ Since Chinese is monowidth, and that column widths are measured by fixed numbers of characters, spacing should be an easy matter. But this is not the case. There are three factors that make justification a complex issue: line-beginning and line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of spacing around certain punctuation marks.
  • 54. Paragraph division: first-line indent (using leading value) Continuous prose in Chinese is customarily set justified left and right. This is true even in classical poetry, where phrases have set number of charac- ters. The fact that Chinese characters are monowidth makes flush left, rag- ged right alignment impractical. Flush left alignment can only be achieved satisfactorily when line breaks are manually forced. As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a surface frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under most circumstances. Column widths (hence line lengths) are calculated by number of characters. Since Chinese is monowidth, and that column widths are measured by fixed numbers of characters, spacing should be an easy matter. But this is not the case. There are three factors that make justification a complex issue: line-beginning and line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of spacing around certain punctuation marks.
  • 55. Paragraph division: line breaks only Continuous prose in Chinese is customarily set justified left and right. This is true even in classical poetry, where phrases have set number of characters. The fact that Chinese characters are monowidth makes flush left, ragged right alignment impractical. Flush left alignment can only be achieved satis- factorily when line breaks are manually forced. As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a surface frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under most circumstances. Column widths (hence line lengths) are calculated by number of characters. Since Chinese is monowidth, and that column widths are measured by fixed numbers of characters, spacing should be an easy matter. But this is not the case. There are three factors that make justification a complex issue: line-beginning and line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of spacing around certain punctuation marks.
  • 56. Paragraph division: line breaks only (flush-left, ragged right) Continuous prose in Chinese is customarily set justified left and right. This is true even in classical poetry, where phrases have set number of charac- ters. The fact that Chinese characters are monowidth makes flush left, rag- ged right alignment impractical. Flush left alignment can only be achieved satisfactorily when line breaks are manually forced. As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a surface frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under most circumstances. Column widths (hence line lengths) are calculated by number of characters. Since Chinese is monowidth, and that column widths are measured by fixed numbers of characters, spacing should be an easy matter. But this is not the case. There are three factors that make justification a complex is- sue: line-beginning and line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of spacing around certain punctuation marks.
  • 57. Paragraph division: hanging indent (or called outdent) Continuous prose in Chinese is customarily set justified left and right. This is true even in classical poetry, where phrases have set number of characters. The fact that Chinese characters are monowidth makes flush left, ragged right alignment impractical. Flush left alignment can only be achieved satis- factorily when line breaks are manually forced. Asmentionedearlier,spacinginChinesetypefacesisachievedinternallywithin each character, by way of defining a surface frame that is a smaller percent- age than the body frame (em square). This internal spacing is the optimal spacing under most circumstances. Column widths (hence line lengths) are calculated by number of characters. Since Chinese is monowidth, and that column widths are measured by fixed numbers of characters, spacing should be an easy matter. But this is not the case. There are three factors that make justification a complex issue: line-beginning and line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of spacing around certain punctuation marks.
  • 58. Paragraph division: line space (full line) Continuous prose in Chinese is customarily set justified left and right. This is true even in classical poetry, where phrases have set number of characters. The fact that Chinese characters are monowidth makes flush left, ragged right alignment impractical. Flush left alignment can only be achieved satis- factorily when line breaks are manually forced. As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a surface frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under most circumstances. Column widths (hence line lengths) are calculated by number of characters. Since Chinese is monowidth, and that column widths are measured by fixed numbers of characters, spacing should be an easy matter. But this is not the case. There are three factors that make justification a complex issue: line-beginning and line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of spacing around certain punctuation marks.
  • 59. Paragraph division: line space (half a line) Continuous prose in Chinese is customarily set justified left and right. This is true even in classical poetry, where phrases have set number of characters. The fact that Chinese characters are monowidth makes flush left, ragged right alignment impractical. Flush left alignment can only be achieved satis- factorily when line breaks are manually forced. As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a surface frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under most circumstances. Column widths (hence line lengths) are calculated by number of characters. Since Chinese is monowidth, and that column widths are measured by fixed numbers of characters, spacing should be an easy matter. But this is not the case. There are three factors that make justification a complex issue: line-beginning and line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of spacing around certain punctuation marks.
  • 60. Paragraph division: line space and first-line indent (redundant) Continuous prose in Chinese is customarily set justified left and right. This is true even in classical poetry, where phrases have set number of characters. The fact that Chinese characters are monowidth makes flush left, ragged right alignment impractical. Flush left alignment can only be achieved satis- factorily when line breaks are manually forced. As mentioned earlier, spacing in Chinese typefaces is achieved internally within each character, by way of defining a surface frame that is a smaller percentage than the body frame (em square). This internal spacing is the optimal spacing under most circumstances. Column widths (hence line lengths) are calculated by number of characters. Since Chinese is monowidth, and that column widths are measured by fixed numbers of characters, spacing should be an easy matter. But this is not the case. There are three factors that make justification a complex issue: line-beginning and line-ending rules; the insertion of proportional width characters, such as Latin glyphs, symbols or figures; adjustments of spacing around certain punctuation marks.