Some will consider my view on the subject (utterly) blurred, while others will consider my posts on Facebook too long by definition. That is why I tried to keep these notes clear and brief.
There is clearly not much exchange of knowledge between scientists
who investigate legibility and type designers. Outcomes of legibility
research seem in general ignored by the latter. Scientists
consider this a mistake. However, it is unlikely that Jenson and Griffo
did any legibility research before they developed their archetypal
models. These two punchcutters, nor any of their Renaissance colleagues,
seem to have investigated the physiological structure of the human
visual system in relation to type. The fact that light falling on the
retina excites photoreceptors was most probably completely unknown to
the early punchcutters. Later in the Renaissance more became known about
the real functioning of the eye. Leonardo Da Vinci researched the
subject and his idea that vision ‘is a result of the eye receiving rays
of light’ was different from the generally accepted idea ‘that humans
had vision because of tiny particles projected from the eye’ (Barbara
O’Connor, Leonardo da Vinci [Minneapolis, 2003] p.51). In 1604 Johannes
Kepler ‘demonstrated the physics behind the optical workings of the eye
as an aside of his astronomical work Supplements to Witelo, On the
Optical part of Astronomy. For the first time, oculists had begun to
accept the retina – and not the lens as it had been assumed since
antiquity – as the organ of vision (Saad Shaikh, Eyes On Ice & No
Blind Mice [Bloomington / Central Milton Keynes, 2007] p.254). There is
no evidence that these outcomes were taken into account by punchcutters.
In his dissertation Legibility, atmosphere-value and forms of
printing types (Leiden, 1938) Gerrit Willem Ovink states that: ‘Every
investigation, which does not stop at stating the superior legibility of
a certain typeface above another, but goes into the causes of such a
fact, ends in framing prescriptions for future typedesigning’ (p.12). In Bases for Effective Reading (Minneapolis, 1966) Miles A. Tinker notes
that the ‘subjective opinions of type designers and typographers as to
legibility of letters prevailed throughout the nineteenth century and
have carried much weight even up to the present day’ (p.125). According
to Tinker type designers actually could (and should) profit from
legibility research: ‘Even though type designers, printers, and
publishers have at times achieved fairly satisfactory legibility by
their approach, application of the results of research would lead to
marked improvement’ (ibid, p.115).
In an interview in 2002
the renowned Dutch type designer and lecturer Gerard Unger stated that
the ‘problem with legibility research is that gives you no clues as a
designer to improve the letterforms’. This statement is clearly in
contradiction with Tinker’s ideas, but I know from experience that in
general type designers don’t read (or are even aware of) the many
publications on legibility, ranging from research on subjects as
‘discrimination of spatially confusable letters by young children’, ‘the
influence of concept training on letter discrimination’, ‘phonology and
reading’, ‘phase of alpha brain waves, reaction time and visually
evoked potentials’, ‘the development of graphemes-phoneme
correspondences’, ‘receptive fields, binocular interaction and
functional architecture in the cat’s visual cortex’ to ‘studies of eye
movements in reading’.
Type designers seem to rely purely on
their eyes, but what they see is the result of conditioning.
Conditioning is based on conventions and conditioning preserves
conventions. Thus the snake bites its own tail; to able to use one’s
‘eye’ like Fournier advocated (Harry Carter, Fournier on typefounding
[New York, 1973] p.9), one has to be educated to look at type in the
same way. Type designers make variants within or outside the borderlines
of the conventions and these conventions are obviously fixed enough,
i.e., provide a safely marked territory, to ignore research such as on
‘visual acuity or eye movement habits’. The fact that ‘serifs are
important in the perception of small letters by humans. They react with
the line detectors of the visual system with the component lines of
letters. The component lines of letters are made easier to see when the
letters are of serif form’ (David Owen Robinson, Michael Abbamonte,
Selby H. Evans ‘Why serifs are Important: the Perception of Small
Print’, Visible Language, Volume V, Number 4, [1971] pp.353–359
[p.356]), may not come up literally into the minds of type designers,
but they are educated in a typographical tradition which is based on
serifed letters. Researches by renowned scientists in the field of
legibility, such as Poulton and Tinker, did not prove that serifed
typefaces are more legible by definition than sans-serif ones. Still
there seems to be a general agreement on the fact that serifed type is
more legible.
Type designers seem to focus especially on what
Ovink called ‘atmosphere-value’. According to Ovink design
(‘form-giving’) in arts and crafts ‘[…] is governed by two systems of
value: beauty versus utility. The latter can be proved objectively, the
former cannot […] the ulterior motives of the artist’s pleading for type
as a reading-tool are of aesthetical nature’ (Legibility,
atmosphere-value […], p.222). But the fact that ‘utility’ is more
objectively measurable than ‘beauty’ does, according to Ovink, not imply
that it is by definition the more important factor: ‘[…] for these
theorists refuse to measure the utility of a type by other qualities
than legibility, even if experiments prove that the reader feels clearly
the atmosphere of a type […]’ (ibid).
What exactly is
legibility? Tinker made a distinction between legibility and
readability: ‘Beginning about 1940, “readability” came to be regarded as
a more descriptive and meaningful word than “legibility,” and it was
commonly adopted. However, with the advent of “readability formulas,”
devised to measure the difficulty of reading material, an entirely
different meaning developed for “readability.” To avoid confusion, it
seems best to employ “legibility of print” to designate the effects of
typographical factors on the ease and efficiency of perception in
reading’ (Bases for Effective Reading, p.115). Nevertheless both terms
are used in a interchangeable manner still: ‘Haultin also used a
‘right-size capital’ together with his lowercase. His capitals were much
smaller and lighter in weight than was then usual, without any loss of
readability’, wrote Fred Smeijers in Counterpunch in 1996 (p.67).
The fact that type designers generally ignore the outcomes of
legibility research does not mean that they don’t have an opinion about
legibility. According to Johnston there are three things, which
constitute legibility: simplicity, distinctiveness, proportion.
Besides these abstract qualities there are various concrete aspect of
legibility; the two most important are accustomedness and fitness’
(Edward Johnston [Heather Child, ed.], Formal Penmanship and other
papers [London, 1971] p.47). His pupil Eric Gill introduced the
difficult to define term ‘beautiful’ in his definition of legibility:
legibility is what the Daily Mail reader finds readable; good style is
what he finds good; the beautiful is what pleases him’ (Eric Gill, An
essay on typography [London, 1988] p.103) According to Goudy beauty does
not by definition guarantee legibility though: ‘Pleasing legibility is
the great desideratum. Beauty, too, is desirable, but beauty must not be
emphasized if it detracts from easy readability. Beauty is an inherent
characteristic of simplicity, dignity, harmony, proportion,
strength—qualities always found in an easily legible type; yet
legibility is seldom achieved by a predetermined effort to produce it’
(Frederic W. Goudy, The Alphabet [New York, 1963] p.91). In line with
Goudy, Tinker notes on the detraction factor: ‘Reader judgements of
relative legibility and pleasingness must also be considered because it
is likely that the criteria of speed and efficiency of performance will
be challenged by those inclined to stress aesthetic values in printing’
(Bases for Effective Reading, p.120).
Actually as an exception to
the rule Goudy seems to have consulted legibility researches; in
Typologia: studies in type design & Type Making [Berkely/London,
1940] he mentions that ‘in 1911 an investigation was undertaken at Clark
University to ascertain “the relative legibility of different faces of
printing types”’ and he subsequently summoned that ‘legibility is a
product of six factors: [1] the form of the letter; [2] the size of the
letters; [3] the heaviness of the face of the letter [the thickness of
the lines which constitute the letter]; [4] the width of the white
margin which surrounds the letter; [5] the position of the letter in the
letter group; [6] the shape and size of the adjacent letters’ (p.142).
Moxon had his own ideas about what made the seventeenth-century Dutch
type legible and, of course, he mentioned the mathematical, i.e.,
geometric, underlying constructions: ‘[…] the late made Dutch-letters
are so generally, and indeed most deservedly accounted the best, as for
their Shape, consisting of Mathematical Regular Figures as aforesaid,
And for the commodious Fatness they have beyond other Letters, which
easing the Eyes in Reading, renders them more Legible […]’ (Joseph Moxon
[Herbert Davis, Harry Carter, ed.], Mechanick Exercises [New York,
1978] p.23). Eric Gill used a simple measurement rule for the point size
in relation to the eyesight: ‘A book held in the hands demands type of
about 10 or 12 point on account of the length of the human arm and the
normal power of human eyesight, assuming a normally legible type’ (An
essay on typography, pp.106,107). Researchers like Tinker use the more
scientific ‘distant method’, ‘which measures visibility or
perceptibility at a distance’ (Bases for Effective Reading, p.117).
Other methods include for instance the measuring of eye-movement and
‘threshold visibility, i.e., the apparatus is manipulated until a word,
letter, or other symbol can just be recognized’ (ibid, p.117).
In the chapter ‘Readers Eyes’ of While you’re Reading (New York, 2007)
Gerard Unger states after describing the fixation of letterforms by the
establishment of the roman type by Jenson and Aldus: ‘The fact that
these forms have since undergone very little fundamental change is not
so much due to Morison’s most conservative reader it is more probably
because these forms had already largely crystallized and were adapted to
the properties of the human eye and brain, i.e. to the ergonomic need
of readers’ (p.116). This is an interesting thought, because it suggests
a superiority of roman type over for instance gothic type. However, one
could argue that the moment in history in which the invention of
movable type took place was coincidentally during the Renaissance and
that for instance any earlier introduction of letterpress printing in
Italy could have had a different result on the fixation of letterforms.
The development of gothic hands since the Carolingian minuscule took
many centuries and hence especially textura was much more crystallized
than the Humanistic letterforms.
The quote from Tinker, in
which he states that although type designers, printers and publishers
have at times achieved fairly satisfactory results but would gain by
applying the results of legibility researches, is not untrue, but Tinker
and his colleagues were and are measuring patterns that are the result
of a refinement process of the written letters and derived type by
respectively scribes and punchcutters. One could state that the
legibility researchers seem to look for absolute values, and maybe there
are only relative ones, which are the results of the applied systems
and models in the graphemes that represent a certain script.
Scientific researches are generally empiric and often make use of
testing groups. This seems to be in line with Tommy Thompson’s statement
that: ‘The eye remains the only machine that can test the design of
letters for it is the only organ that can read them’ (Tommy Thompson,
How to render roman letter forms (new York, 1946) p.30). However, if there
really are objective criteria for measuring legibility, perhaps a machine,
i.e., software, should do the measurement? This would exclude
unclear subjective preferences, like taste and the perception of beauty,
or at least would make it possible to map these factors, described by
Ovink as ‘atmosphere value’, against generic models. This requires a
clear mapping of the aspects and particles involved. I made a brief
start here and here.
Comments
Cross-post here, too.
Gentlemen, your wish is my command!
There is clearly not much exchange of knowledge between scientists who investigate legibility and type designers. Outcomes of legibility research seem in general ignored by the latter. Scientists consider this a mistake. However, it is unlikely that Jenson and Griffo did any legibility research before they developed their archetypal models. These two punchcutters, nor any of their Renaissance colleagues, seem to have investigated the physiological structure of the human visual system in relation to type. The fact that light falling on the retina excites photoreceptors was most probably completely unknown to the early punchcutters. Later in the Renaissance more became known about the real functioning of the eye. Leonardo Da Vinci researched the subject and his idea that vision ‘is a result of the eye receiving rays of light’ was different from the generally accepted idea ‘that humans had vision because of tiny particles projected from the eye’ (Barbara O’Connor, Leonardo da Vinci [Minneapolis, 2003] p.51). In 1604 Johannes Kepler ‘demonstrated the physics behind the optical workings of the eye as an aside of his astronomical work Supplements to Witelo, On the Optical part of Astronomy. For the first time, oculists had begun to accept the retina – and not the lens as it had been assumed since antiquity – as the organ of vision (Saad Shaikh, Eyes On Ice & No Blind Mice [Bloomington / Central Milton Keynes, 2007] p.254). There is no evidence that these outcomes were taken into account by punchcutters.
In his dissertation Legibility, atmosphere-value and forms of printing types (Leiden, 1938) Gerrit Willem Ovink states that: ‘Every investigation, which does not stop at stating the superior legibility of a certain typeface above another, but goes into the causes of such a fact, ends in framing prescriptions for future typedesigning’ (p.12). In Bases for Effective Reading (Minneapolis, 1966) Miles A. Tinker notes that the ‘subjective opinions of type designers and typographers as to legibility of letters prevailed throughout the nineteenth century and have carried much weight even up to the present day’ (p.125). According to Tinker type designers actually could (and should) profit from legibility research: ‘Even though type designers, printers, and publishers have at times achieved fairly satisfactory legibility by their approach, application of the results of research would lead to marked improvement’ (ibid, p.115).
In an interview in 2002 the renowned Dutch type designer and lecturer Gerard Unger stated that the ‘problem with legibility research is that gives you no clues as a designer to improve the letterforms’. This statement is clearly in contradiction with Tinker’s ideas, but I know from experience that in general type designers don’t read (or are even aware of) the many publications on legibility, ranging from research on subjects as ‘discrimination of spatially confusable letters by young children’, ‘the influence of concept training on letter discrimination’, ‘phonology and reading’, ‘phase of alpha brain waves, reaction time and visually evoked potentials’, ‘the development of graphemes-phoneme correspondences’, ‘receptive fields, binocular interaction and functional architecture in the cat’s visual cortex’ to ‘studies of eye movements in reading’.
Type designers seem to rely purely on their eyes, but what they see is the result of conditioning. Conditioning is based on conventions and conditioning preserves conventions. Thus the snake bites its own tail; to able to use one’s ‘eye’ like Fournier advocated (Harry Carter, Fournier on typefounding [New York, 1973] p.9), one has to be educated to look at type in the same way. Type designers make variants within or outside the borderlines of the conventions and these conventions are obviously fixed enough, i.e., provide a safely marked territory, to ignore research such as on ‘visual acuity or eye movement habits’. The fact that ‘serifs are important in the perception of small letters by humans. They react with the line detectors of the visual system with the component lines of letters. The component lines of letters are made easier to see when the letters are of serif form’ (David Owen Robinson, Michael Abbamonte, Selby H. Evans ‘Why serifs are Important: the Perception of Small Print’, Visible Language, Volume V, Number 4, [1971] pp.353–359 [p.356]), may not come up literally into the minds of type designers, but they are educated in a typographical tradition which is based on serifed letters. Researches by renowned scientists in the field of legibility, such as Poulton and Tinker, did not prove that serifed typefaces are more legible by definition than sans-serif ones. Still there seems to be a general agreement on the fact that serifed type is more legible.
Type designers seem to focus especially on what Ovink called ‘atmosphere-value’. According to Ovink design (‘form-giving’) in arts and crafts ‘[…] is governed by two systems of value: beauty versus utility. The latter can be proved objectively, the former cannot […] the ulterior motives of the artist’s pleading for type as a reading-tool are of aesthetical nature’ (Legibility, atmosphere-value […], p.222). But the fact that ‘utility’ is more objectively measurable than ‘beauty’ does, according to Ovink, not imply that it is by definition the more important factor: ‘[…] for these theorists refuse to measure the utility of a type by other qualities than legibility, even if experiments prove that the reader feels clearly the atmosphere of a type […]’ (ibid).
What exactly is legibility? Tinker made a distinction between legibility and readability: ‘Beginning about 1940, “readability” came to be regarded as a more descriptive and meaningful word than “legibility,” and it was commonly adopted. However, with the advent of “readability formulas,” devised to measure the difficulty of reading material, an entirely different meaning developed for “readability.” To avoid confusion, it seems best to employ “legibility of print” to designate the effects of typographical factors on the ease and efficiency of perception in reading’ (Bases for Effective Reading, p.115). Nevertheless both terms are used in a interchangeable manner still: ‘Haultin also used a ‘right-size capital’ together with his lowercase. His capitals were much smaller and lighter in weight than was then usual, without any loss of readability’, wrote Fred Smeijers in Counterpunch in 1996 (p.67).
The fact that type designers generally ignore the outcomes of legibility research does not mean that they don’t have an opinion about legibility. According to Johnston there are three things, which constitute legibility: simplicity, distinctiveness, proportion. Besides these abstract qualities there are various concrete aspect of legibility; the two most important are accustomedness and fitness’ (Edward Johnston [Heather Child, ed.], Formal Penmanship and other papers [London, 1971] p.47). His pupil Eric Gill introduced the difficult to define term ‘beautiful’ in his definition of legibility: legibility is what the Daily Mail reader finds readable; good style is what he finds good; the beautiful is what pleases him’ (Eric Gill, An essay on typography [London, 1988] p.103) According to Goudy beauty does not by definition guarantee legibility though: ‘Pleasing legibility is the great desideratum. Beauty, too, is desirable, but beauty must not be emphasized if it detracts from easy readability. Beauty is an inherent characteristic of simplicity, dignity, harmony, proportion, strength—qualities always found in an easily legible type; yet legibility is seldom achieved by a predetermined effort to produce it’ (Frederic W. Goudy, The Alphabet [New York, 1963] p.91). In line with Goudy, Tinker notes on the detraction factor: ‘Reader judgements of relative legibility and pleasingness must also be considered because it is likely that the criteria of speed and efficiency of performance will be challenged by those inclined to stress aesthetic values in printing’ (Bases for Effective Reading, p.120).
Actually as an exception to the rule Goudy seems to have consulted legibility researches; in Typologia: studies in type design & Type Making [Berkely/London, 1940] he mentions that ‘in 1911 an investigation was undertaken at Clark University to ascertain “the relative legibility of different faces of printing types”’ and he subsequently summoned that ‘legibility is a product of six factors: [1] the form of the letter; [2] the size of the letters; [3] the heaviness of the face of the letter [the thickness of the lines which constitute the letter]; [4] the width of the white margin which surrounds the letter; [5] the position of the letter in the letter group; [6] the shape and size of the adjacent letters’ (p.142). Moxon had his own ideas about what made the seventeenth-century Dutch type legible and, of course, he mentioned the mathematical, i.e., geometric, underlying constructions: ‘[…] the late made Dutch-letters are so generally, and indeed most deservedly accounted the best, as for their Shape, consisting of Mathematical Regular Figures as aforesaid, And for the commodious Fatness they have beyond other Letters, which easing the Eyes in Reading, renders them more Legible […]’ (Joseph Moxon [Herbert Davis, Harry Carter, ed.], Mechanick Exercises [New York, 1978] p.23). Eric Gill used a simple measurement rule for the point size in relation to the eyesight: ‘A book held in the hands demands type of about 10 or 12 point on account of the length of the human arm and the normal power of human eyesight, assuming a normally legible type’ (An essay on typography, pp.106,107). Researchers like Tinker use the more scientific ‘distant method’, ‘which measures visibility or perceptibility at a distance’ (Bases for Effective Reading, p.117). Other methods include for instance the measuring of eye-movement and ‘threshold visibility, i.e., the apparatus is manipulated until a word, letter, or other symbol can just be recognized’ (ibid, p.117).
In the chapter ‘Readers Eyes’ of While you’re Reading (New York, 2007) Gerard Unger states after describing the fixation of letterforms by the establishment of the roman type by Jenson and Aldus: ‘The fact that these forms have since undergone very little fundamental change is not so much due to Morison’s most conservative reader it is more probably because these forms had already largely crystallized and were adapted to the properties of the human eye and brain, i.e. to the ergonomic need of readers’ (p.116). This is an interesting thought, because it suggests a superiority of roman type over for instance gothic type. However, one could argue that the moment in history in which the invention of movable type took place was coincidentally during the Renaissance and that for instance any earlier introduction of letterpress printing in Italy could have had a different result on the fixation of letterforms. The development of gothic hands since the Carolingian minuscule took many centuries and hence especially textura was much more crystallized than the Humanistic letterforms.
The quote from Tinker, in which he states that although type designers, printers and publishers have at times achieved fairly satisfactory results but would gain by applying the results of legibility researches, is not untrue, but Tinker and his colleagues were and are measuring patterns that are the result of a refinement process of the written letters and derived type by respectively scribes and punchcutters. One could state that the legibility researchers seem to look for absolute values, and maybe there are only relative ones, which are the results of the applied systems and models in the graphemes that represent a certain script.
Scientific researches are generally empiric and often make use of testing groups. This seems to be in line with Tommy Thompson’s statement that: ‘The eye remains the only machine that can test the design of letters for it is the only organ that can read them’ (Tommy Thompson, How to render roman letter forms (new York, 1946) p.30). However, if there really are objective criteria for measuring legibility, perhaps a machine, i.e., software, should do the measurement? This would exclude unclear subjective preferences, like taste and the perception of beauty, or at least would make it possible to map these factors, described by Ovink as ‘atmosphere value’, against generic models. This requires a clear mapping of the aspects and particles involved. I made a brief start here and here.