You are absolutely right. When type becomes too large, the number of words on each line becomes too small. This means that the eye has to spend a lot of time jumping from line-end to line-start instead of actually reading. Also, almost all (i can't think of any who don't) typefaces have a limited range of size in which they work well. Typefaces for body text usually don't work well when set too large, they tend to become unelegant and too low-contrast (when using a serif typeface). The opposite is the case when setting titling typefaces too small – the small details become too small and may disappear in print.

But when it comes to determining the "ultimate body copy" size, that's just impossible. If you line up a variety of typefaces in the same size, say 11pt, they will all look different in size due to their different x-heights.

good luck on convincing your editors :-)

alex

What font will you use, I attach a sample that shows two typefaces in the same body size.

I think the most important point (!) is that point size alone means very little.

First of all x-height tells you more about legibility, as mentioned. But also length of line and leading makes a crucial difference. Also sans vs serif.

I know some people get in their heads that point size is everything. I think the only way to disabuse them of that idea is to show them several side-by-side versions of type actually set in a paragraph.

That being said I think that the most common mistake of those who don't work with type is that "Make It Bigger" (title of Paula Sher's book) solves all problems. On the other hand, I do think that graphic designers sometimes make the opposite mistake. Because they want the overall page to look nice, they make the type too small or low contrast (colors, reversed, etc.) for good readability. You can definitely sin in both directions.

There is an advice from RNIB which claims that 12 pt is easier for the vision-impaired to read. However despite this claim "RNIB’s clear print guidelines are based on our experience of the issues over many years, advice from experts in the field and evidence including recent research into fonts and type size." it doesn't say what the recent research was or who the experts in the field were, unfortunately many clients just seem to read the 12pt bit and don't bother to consider the rest, like, for example, the part about the x-height being 2–2.3mm high.
http://www.rnib.org.uk/xpedio/groups/public/documents/publicwebsite/publ...

Tim

I've been wondering about this topic again, mostly because I'm reading "The Grid." Does anyone know why Hurlburt set the text in the book so large (It looks like 14 points)?

What is the set width? Depending on that value, 12pt might be just fine and 8pt too small.

Point size is an increasingly meaningless number, x-height and cap-height are more important.

Who is the audience? If it is for persons 45 and older, it better be a little larger than a touch too small.

Just remember one person's horsey type is another persons minimum requirement.

I wish I could start this thread over and re-phrase my question.

Yes, point size is a meaningless number. Let's be specific-- I'm currently working with Adobe Garamond pro on a book about Napier New Zealand-- I was told to use 12 point by editors-- but it looked huge and awkward, no room for white space, everything looked cramped on the A4 size pages--

I also have to put in photos and captions. When I reduce the point size down to 10 on 13 Garamond-- the two page spread looks so much better. (9pt looks even better)

As for audience -- it's not the visually impaired, or the elderly, or children-- its meant for a very general audience ages 20-40.

So, if you want to do a test--- a typical spread (two A4 size pages) in this book has 1100-1300 words, with 4 photos with captions 40-60 words. There is a title, subtitle. Try designing a good looking spread (two A4 pages)-- with 12 point Garamond-- then do it with 10 pt -- see which one looks horsey. See which one is easier to read.
(What you are going to find is that on spreads with 1300 words with 12 pt type you will have to reduce the leading to get all the words on the page- it begins to look amateurish).

But aesthetics aside....
-- The whole point of my post-- really, is to find out if there has been any study that body copy set to match the size of 12 pt garamond is not easier to read than type set to the size of 9 or 10 pt Adobe Garamond pro. (for people who are not vision impaired, over the age of 60 or under the age of 8.)

I would love to know if there is a
test similar to the test on this site (which is on legibility of web fonts)
on type size.
http://psychology.wichita.edu/surl/usabilitynews/81/legibility.htm

I'm going to grossly misquote Gill and say that "familiarity is everything." Europeans don't seem to have nearly as much trouble with sans type, because they're used to it. People who only read cheap paperback novels don't have any trouble with eight-point type. Those with bad vision have no trouble with 14+ point type because its all they ever see. Many of my friends are lawyers, and they whinge about reading anything that isn't set in 12 point Times New Roman with with double-spaced lines and extra spaces after periods.

Don't worry about studies or rules for type size. Worry about what your audience is expecting and how you can use their expectations to manipulate them.

You're not the first to ask for the results of "readability" tests to satisfy a client (or oneself). it must be one of the most-asked questions on Typophile. Why do typographers and their clients need to be reassured by some banal tests made by scientists who have no feeling for typography that they know what they're doing? We all know the tests are meaningless (note Alessandro's sample). Whose eyes are we pulling the wool over? The more this goes on, the more we dig our own grave. If we let our clients (or ourselves) buy into this bullshit that the the quality of typography can be measured, they won't need typographers, but will get everything set in Arial or Comic Sans (whatever is "proved" most legible) in the "optimum" setting. Why do we need authorities to justify our decisions? Isn't the designer the authority? What is the point of developing a refined sensibility if no-one trusts it? If that's the case, the typographer is a low-grade technician who can soon be replaced by a piece of software.

However, why not prepare a few variants, and test them on a suitable focus group? That's much better than relying on vague generalization.

Well said, Nick!

I don't think what Nick said was well said at all. Is relying on a typographers opinion better than relying on a well conducted study? Nick asked, "Why do typographers and their clients need to be reassured by "banal" tests?" Noone wants to be reassured by a "banal" test— they want to be reassured by well conducted test. Can there be a valid test?- not if the attitude is that all tests are banal. Nick says the tests are all meaningless— and uses Allesandro's test as an example-- If thats Nick's idea of a test-- then no wonder you think tests are banal.

"anyone know why Hurlburt set the text in the book so large"

He needed to fill the book?

ChrisL

Jasper, I don't think that all tests are bad or that advancing our knowlege of type through science is a bad idea; but as far as I have seen ( which not far enough I will admit ) there is a massive massive gap between what has been done with testing so far & what type designers ( or type users ) can actualy find broadly useful. My take on what Nick said is not 'screw science' - it's that there is no point in waiting around with our hands in our pockets for science to catch up and guide us. As he said ' What is the point of developing a refined sensibility if no-one trusts it? '. He is bang on there. Plenty of type designer do know what they are doing. And how do they do that great work? A: They use that sensibility that Nick mentioned. When you can point to a well desined and typographically sensible test or two, by all means let us know! But don't expect us to hold our breath in the meantime.

Hi Jasper, my name is with one 'l' and two 's', don’t put four dots after “aesthetics aside”, my sample is not Nick’s idea of a test if I can speak for him, I think what he wrote rocks, and why you don’t question the A4 format for example, or why Adobe Garamond and not ITC. If you are interested I have got some links on legibility here : http://www.as8.it/edu/bibliography-TD.html

Nick wrote:
“Why do we need authorities to justify our decisions? Isn’t the designer the authority?”

Nick, I think the reader is the authority. For physiological, emotional, and experiential reasons, some designs are better than others. One designer may choose to use 9/10 Garamond while another uses 10/12. If the goal is to make it as quick as possible for readers to read this information, can both designs be equally optimal?

Reading tests shouldn’t be banal or used to justify anyone’s decisions, but rather to learn more about how people read. This knowledge can be used to improve designs. Test do not replace the role of the designer in any way. Testing can only be used to study existing reading conditions, but improving the experience beyond what already exists takes creativity. All design involves creative problem solving which cannot be automated.

Miles Tinker ran many studies investigating optimal type sizes. For most typefaces he found that reading speed peaked at 11 point and was slower when you used smaller or larger text. For some fonts (I believe those with larger x-heights) reading speed peaked at 10 point. See Tinker’s book Legibility of Print for more details.

This assumes a reader with normal visual abilities. With some vision impairments, it is not possible to read 8-12 point text sizes and larger text is required. While reading is impossible without a much larger size, it does accompany a loss in reading speed. Optimal text size is a natural consequence of the density of photoreceptors in the fovea (part of the retina used for reading).

Cheers, Kevin

You know, this design does a great job of reminding me of the one thing I really like about designing in XHTML/CSS. As long as I create an accommodating layout and don't set the font size in pixels, the user can just increase or decrease the font size to to whatever makes him comfortable. That's just so awesome it hurts.

Kevin,
Tinkers studies are not very recent. The nature of both printing, type design, have gone through several changes since then. What readers today have gotten accustomed to reading has also changed. There is a lot more smaller type being read than their used to be. I don't pretend to know if any of this changes Tinker's conclusions for today's reader but I sure would like someone to test it.
On another readability issue, much has been made of eye movements as a measure of good readability but I have a problem in the translation from percent of difference of eye movement to percent of enhanced or decreased readability. I have yet to see a comparative measure which gives me a usable comparison in reading speed and reading comprehension. An example might be using today's Garamond Premier Pro at 9 point vs. 10 point with the same characters per line and same proportion of leading, how much faster in pages per hour and how much more comprehension (however you can really measure that) am I gaining or losing by changing the point size one point?
This will tell me something to help make design decisions. If the difference is so small that it is inconsequential in a given reading situation then why not opt for the more efficient use of space? Readability is always paired with efficiency in the decision making process. The readability that interests me is when it affects the exact situation it is being used in, not just a more generalized "bigger is better" or "serif is better" thinking which gets seized upon rather than what the actual studies conclusion intended.

ChrisL

>For most typefaces he found that reading speed peaked at 11 point and was slower when you used smaller or larger text.

That agrees with my intuitive feeling, if we take an average ascender to be 1.6 bigger than the x height, and the type to fill 95% + of the body size.

Kevin, I think that you could refine these numbers if you would use my suggested measure of reading comfort: rate of decrease in comprehension with time :). You could also measure boldness (% of black to white), presence of serifs and amount of leading--all of which I think will have a measureable influence. And letter spacing and measure. Those are 6 variables, and though a lot would tell us most of what we need to know.

For what it's worth, I do think that below 11 point Adobe Garamond Pro is less readable, as it has a small x-height. However, it will probably require less leading at 11 or 12 point, so it may end up being economical enough for you.

In one ear, out the other.

hhp

" Is relying on a typographers opinion better than relying on a well conducted study?"
Opinion of What? Study of what? In my opinion, if you have both pertaining to your problem, it's best, (especially if they agree:). But if you have to choose between typographic practice, and a study of something else, other than what you need a study of, well, I'd go with the practice. And on the other hand, if I had a study of what I was looking for, and typographic practice of something else, well, then, I'd go read the study carefully to make sure, because one bad move in a study can muddy the results beyond belief, so to speak.

"Does anyone know of any studies [of] legibility that show...Body copy larger than 9, 10, 11 pt [] are actually harder to read? (I prefer reading text that is between 8-10 pt)"
We did a preference study at the Poynter Institute more than 10 years ago that showed readers preferred type set in newspapers that was designed for setting between 8.5 and 11 pt., and set at those sizes, not smaller or larger. Not all of the types we tested were designed for such use, and those did not win, but they did better than expected. All of the subjects were expecting "newspaper text", (the paper being a dead giveaway), and they were all adults, so it wasn't a test of ALL LEGIBILITY, or even all fonts, or all users.

But the results allowed us to see how to draw some very readable faces, for print, that if sales are any indication, which I'm sure they are not, are very readable, i.e. beyond legible perhaps. We are planning now to do the same thing for screen types, building a test-bed of fonts, and sending them home with users to try out. I'm arguing for both obtrusive and non-obtrusive recording of the choices, but we'll see. It's all very expensive, and you need to get it right from initial input to the final adjustments of claims. Just like typeface design, if everything isn't right, it useless.

http://www.designwritingresearch.org/essays/science.html

"Poynter Institute more than 10 years ago that showed readers preferred type set in newspapers that was designed for setting between 8.5 and 11 pt., and set at those sizes, not smaller or larger."

That is intriguing and is what I would have thought modern day readers would think. Is this study available to us?

ChrisL

In connection with the research on newspaper types that David cites, I think it important context that they usually have large x heights and are set in narrow columns. This might explain why the sizes below 11 point are good also, an apparent contradiction to the study Kevin cites.

The Tinker studies were begun in the 1930s I believe. There has been a lot of water over the dam since then.

ChrisL

As one with a little experience in book design for the New Zealand market, I feel compelled to say that the designer's 'authority' is (unfortunately) a moot point in this context.

Your editors, Jasper, are telling you to use the 12 pt Garamond because they've previously had a good experience with that particular spec. and they're just going with what they know. If you try and railroad them into accepting your preference - with or without some other authority - it's likely to be counterproductive.

You don't mention the other attributes of the spec. but notice how everyone else does; without the associated values that make up the rest of the appearance of the text block, arguing about 12pt or not is a little abstract.

I can suggest three possible approaches;

1. Offer them options; do a truckload of comparison settings (whole pages) with variations of line length (set width), leading and tracking. Dig out the rule-of-thumb reminders about optimum line length and leading values. Be prepared to make minute variations and annotate everything so you know at a glance what the variables are. Don't forget to consider paragraph indentation and/or spacing and how many lines you'll achieve for each page. Sit down with the editors and discuss the results as openly as possible.

2. Do a word count and work out the average page yield. Calculate the overall extent of the book at 12pt. Do the same calculation based on your preference for the smaller type size; then try doing it again for an interim value between the two. Try and figure out if it makes a difference in terms of the print production costs associated with the book.

3. Research other titles on Napier, preferably by competitor publishing houses here in NZ. Are they using 12pt Garamond too? Try and find examples of whatever won last years NZ book awards. What will be this title's point of difference from all the other books?

Good luck with it.

Nice Post Ben! You have taken a pracatical approach rather than a theoretical one and also pushed Jasper to look for himself ( and to get his editors to look as well ). Even for an expert (who aught to be respected) collaboration is where it's at. And collaboration; if you prepare for it, can foster that respect.

ChrisL wrote:
> Tinkers studies are not very recent.

Tinker’s legibility research program was huge, carried out over 4 decades (1920s through the 1950s) on many thousands of readers. It would be great if someone would undertake the project to retest his most famous findings to see if anything has changed since Tinker’s time. My guess is that the text size finding hasn’t changed because it’s a result of photoreceptor density, not personal experience. In one of the Tinker & Patterson books they reports on a survey of typographic variables like point size and leading used in different kinds of publications (novels, newspapers, textbooks, etc.). I was surprised at how common it was to see text sizes as small as 6 point.

ChrisL wrote:
> On another readability issue, much has been made of eye movements as a
> measure of good readability but I have a problem in the translation from percent
> of difference of eye movement to percent of enhanced or decreased readability.

Eye tracking measures are similar to reading speed measures, but provide additional detail such as the amount of time spent fixating on each word and the frequency of regressive saccades. Most studies look for a difference either in reading speed or comprehension. When measuring reading speed, the material used is simple enough that readers will understand it with full comprehension and comprehension questions are used to confirm this. Else more difficult material is used and the amount of time spent reading is controlled (e.g. you are given 5 minutes to read an article). This style of reading test is frequently used in U.S. college admission tests. Combinations of reading speed and comprehension are rarely used because they the scores need to be combined, and debated if it better to read an article in 4 minutes with 80% comprehension or 5 minutes with 90% accuracy?

David wrote:
> We did a preference study at the Poynter Institute more than 10 years ago that
> showed readers preferred type set in newspapers that was designed for setting
> between 8.5 and 11 pt., and set at those sizes, not smaller or larger.

I’m not necessarily surprised that people said they preferred smaller text sizes for newspapers. It wouldn’t be unreasonable at all for people to desire a smaller text size than the one that allows for fast reading. Many people read newspapers by only reading the front page of each section. By asking for smaller text sizes, the readers asking for more information on that first page. I don’t see this as being in conflict with the reading speed data.

Cheers, Kevin

Not too long ago I also had a similar problem. The director was convinced that sans was more legible and that bigger was better. I preferred my 9pt Warnock. So, I obliged her with 3 spreads in each. She physically recoiled from the squishy sans layout and agreed wholeheartedly to my Warnock. I compromised and went up to 9.5pt.

(victory!)

K.

Nick, I think the reader is the authority.

Kevin, you're mistaking the user's degree of satisfaction with a professional service for actual professional expertise in providing satisfaction.

For instance, you may have a subjective opinion of whether a doctor's treatment has effected your health, but this does not make you an authority on medicine.

Optimal text size is a natural consequence of the density of photoreceptors in the fovea

That doesn't even make sense. There is no such thing as optimal text size, because there is no way to measure text size. It is quite simply not a scientific concept. The best that can be determined is that easy-reading text should be between 8 and 12 pt size, which is so vague and useless a statement as to be a self-evident platitude, like saying gloves should not be too tight or too loose.

It is not physiology which best determines the text size for a particular setting, but a typographic understanding of the many variables involved in reading it, such as the reader demographic, the paper stock, the publishing genre (which affects the reading environment), the document page size, language, author, paragraph length, and of course the many typographic variables other than size, such as typeface, leading, measure (line length), tracking, alignment style, gutter and margin width.

>Kevin, you’re mistaking the user’s degree of satisfaction with a professional service for actual professional expertise in providing satisfaction.

Nick, your analogy is not sound. Kevin reports that "reading speed" peaked at 11 point. This is not a subjective self-report of satisfaction, but an objective measure. In science, the results of objective tests are the authority. They trump scientists' opinions and your opinion and mine.

A closer analogy would be to Doctors' opinions of the effectiveness of a procedure and the results of objective and extensive double-blind tests of the actual impact on patients. The results are more authoritative than the opinions. Of course there can be bias, poor methodology, and cheating in the tests. But these can be checked for by repeating experiments.

Time to "repeat the experiments" using modern type and conditions. Tinker tested Kabel against Garamond #3, I think we have come a long way since then.

ChrisL

At one point Science told us that the World was flat and that Autism was caused by cold-unloving mothers. Later experiments proved these to be wrong.

ChrisL

>Science told us

What scientific research actually shows and what is claimed in the name of science are often very far apart, particularly in the fields of medicine and health.

Drawing conclusions from experimentation is not easy unless there are very well articulated theories and great tests.

>Time to “repeat the experiments” using modern type and conditions.

By all means, new tests. However, in order to justify new experiments, I am sure that Kevin or another researcher would want to have a strong reason to expect different results. What is different about modern and type and conditions that would lead us to expect a different result? Or what was defective in the methodology of the earlier research?

As I wrote, a new measure--comfort in reading, as measured by decrease in comprehension with time--might reveal a lot of information that reading speed alone doesn't tell us. So I would like the 6 variables I mentioned tested this way, as well as type size--or preferably, x height size.

Nick, your analogy is not sound. Kevin reports that “reading speed” peaked at 11 point. This is not a subjective self-report of satisfaction, but an objective measure.

So the the reader is the authority by dint of performing in a reading speed test? That's even more ludicrous.

In science, the results of objective tests are the authority.

And not the subjects of the test. So you agree with me.

"Don't you touch my Art, bitch!"

hhp

I agree completely with William’s comments about objective measures. Repeatable objective measures are more interesting that individual opinion. As Chris points out, it was once people’s subjective opinion that that the world was flat. But it’s difficult to rectify this theory with objective data that the sun rises earlier and falls earlier in New York than Los Angeles. I don’t think there is any objective data that supports that flat world theory.

I think there are other possibilities for measuring the reading experience, including William’s proposal and a measure that I proposed at ATypI Helsinki (which will be published soon); but I would claim that reading speed is the primary measure of legibility. If I’m given the choice between reading an article with two typographic treatments and told that I’ll read one version 5% faster without a loss in comprehension, I’m going to choose that version.

Cheers, Kevin

"I don’t think there is any objective data that supports that flat world theory"

Of course not now. What was thought to be objective data centuries ago was that sometimes ships went out to sea and never returned. The falacious but nevertheless agreed upon wisdom of that time was that they fell off the edge of the Earth. My point was that what we may now think of as objective data may at some point in the future be thought of as wrong. This will never occur if we don't question past measurements or view things from "outside the box". I am just saying don't lock up our minds with absolutes so much that we never look for things we were unable to see before. Is Pluto a planet now? Was it thought to be a planet last year? Did the astronomers quit doing science because they had "objective data" that pluto was a planet? No, they kept looking. Let's keep looking too.

Kevin: Where does the "5% faster" figure come from or are you just pulling a number out of the air for sake of argument?

ChrisL

“Don’t you touch my Art, bitch!”

Well, if they want to discredit my expertise as "personal opinion" they'll have to do a more convincing job of it. Let's see those "repeatable objective measures" in action! However, I suspect that the description of the experiment would have to be so typographically precise, in order to get the same results, that generalization would be impossible. If Scientist A uses Adobe Garamond and Scientist B uses ITC Garamond...

As Chris points out, it was once people’s subjective opinion that that the world was flat.

And it is still people's subjective experience that the sun rises and sets. If you imagine the earth spinning, it makes for dizzyness.

Deaf & Deafer.

hhp

Just don't show them your handwriting, or let them measure your cranium.

"At one point Science told us that the World was flat..."
and then it took a while but they determined that is was round. But then they proved it was a sphere, and then later an ovoid shape and now, it's a shrinking ovoid shape.

“reading speed” peaked at 11 point. This is not a subjective self-report of satisfaction, but an objective measure."
But it is an objective measure only of 11 point plus, a long string of other defined variables from illumintation to type style. It could easily be repeated with a slightly younger average age and 10.5, if available, would win. I think, that it's only a truely objective measure when you know exactly what will make 10.5 win, or 11 lose. It's a truely valuable objective measure when all the variables have been published for further research to build on, variable by variable.

>an objective measure only of 11 point plus, a long string of other defined variables from illumintation to type style.

How best to interpret the past results is open to question, but it sounds to me like they are significant, though they need a lot of refinement. I agree about the need for further variables--I mentioned above six or seven to be tested.

If I go to a baker & tell him that I have proved with some sort of test that 'Brioche is the best bread & here are my 3 variables I tested on' I will be laughed at. And rightly so.

Bakers don't need scientist to tell them how to make bread. Of course Science can be used to find out lots of interesting things about bread. But before a scientist studies bread they should get to know bread from the Bakers point of view. As far as I understand it that 'getting to know' the subject hasn't happened with type studies - yet. The tests have been too naive. MUCH too naive. That could change and I would like it to change. But until it does there is is no point in waiving 'science' & 'tests' around to impress us. A little more meekness would go a long way. Nick's examples of variables to think about are excellent. No doubt there are even more to consider.

In the end it never going to be a good idea to ask 'tests' what bread you should eat or what type you should use. Rather, it's always going to be a good idea to feel your way into it, thinking about the situation as whole, and experiemening for yourself, and when you can; allowing your decision to be informed by whatever science can offer at the time. These choices are always going be partly a cultural excercise too. The pure consideration of efficiency and or effectiveness will never be a complete guide to practice.

I think that people who are beating down these proclimations of 'fact' are not against science at all. I think they were interested and then disappointed by what they found. Maybe I am just gerealizing my own case.

Ceratinly there are always people who rather than working at looking & seeing better ( for themselves ) would like to rely on someone else's rules to justify what their choices. Designers and Clients. But that's just hiding from reality. It's weak. No matter what science may get around to showing in time it's always going to be a good idea to keep your eyes open & mind open. That's just good science.

I am not going to to go so far as to say I'm right about any of this. But this how it looks to me today.

The Earth is flat myth is from the 19th century, from Aristotle (4th century BC) and right through the Middle Ages it was known the world is roughly spherical, those that claimed otherwise were the exception rather than the rule.
Apologies for the off-topic, but since we are talking of science here.

Tim

Eban Sorkin wrote:
>In the end it never going to be a good idea to ask ‘tests’ what bread
>you should eat or what type you should use.

Agreed. But as you point out, science is being quite extensively by cooks. Many cooking magazines, led by Cook’s Illustrated, regularly use the scientific method to improve their recipes. They do not test if brioche the best bread, but rather attempt to optimize the preparation of a certain recipe. They regularly make 20 different versions of a recipe, manipulating one variable at a time, in their search for the best preparation. They ask several people to taste each version of the recipe and rate how good a recipe tastes. Testing is being used here as a useful tool to help the chef.

At a different level, Harold McGee has written several interesting books explaining how different cooking techniques impact foods at the molecular level. Our understanding of the effects of cooking preparations on food is better than our understanding of legibility (though certainly not perfect). This information is useful for understanding the health impact of different foods and preparations. If a chef is interested in making healthful food, it would be useful to pay attention to this kind of information. Similarly, if we understood the mental processes for recognizing words, this information could be used to make more legible typefaces. Of course if a type designers purpose had nothing to do with legibility, this information could be safely ignored.

There has been a misunderstanding about science in this thread. Science is not a topic, but a method for learning. Not all statements are scientific just because they are about topics closely associated with science like chemistry, physics, or the natural world. Science is defined by the scientific method, a process for making observations about the world and conducting experiments to test the accuracy of proposed theories. Claiming that ships could go missing because they fell off the edge of the world is not a scientific statement, but a post-hoc explanation of an event. A repeatable scientific experiment that could support the flat earth theory would be to demonstrate that the sun rises and falls at the same time in Seattle and Los Angeles. The best theory is the one that supports the most evidence.

Cheers, Kevin

Returning to the topic of type size, I thought it might be valuable to print the Table of Contents from Miles Tinker’s book Legibility of Print:

1. Introduction
2. Methodology and Definitions
3. Legibility of Letters and Digits
4. Kinds of Type
5. Size of Type
6. Width of Line
7. Leading and Relationship of Leading, Type Size, and Line Width
8. Spatial Arrangement of the Printed Page
9. Color of Print and Background
10. Printing Surfaces
11. Cumulative Effect of Combining Nonoptimal Typographical Arrangements
12. Newspaper Typography
13. Formulas and Mathematical Tables
14. Special Printing Situations
15. Illumination for Reading
16. The Hygienic Reading Situation

I am quite impressed by the breadth of topics that Tinker studied, and am surprised that many typographers are not familiar with his work. It has been mentioned in this thread that some researchers understand little about typography. This does not apply to Tinker. He spent his entire career studying typography and working with printers.

In this book you can find studies of size effects on different type faces in Chapter 5, studies looking at the interaction of leading line length and size in Chapter 7, studies of size in the context of Newspapers in Chapter 12, and the interaction of size and illumination in Chapter 15.

Cheers, Kevin

I have heard that if anybody can make progress in this area it's you Kevin. And I agree with your statement here and wish you and your type projects all possible success and godspeed! I don't doubt for a moment that well designed tests could help type design in many many and possibly profound ways. If you agree that Type Design and Type Use are however, always going be partly cultural excercises too in which efficiency and or test-proven effectiveness will never be a complete guide to practice; why then we would seem to agree completely! Actually, I would love to chat with you sometime. Is that possible?

> I have heard that if anybody can make progress in this area it’s you Kevin.

I told you that. :-)

hhp

Indeed you did.

Hi Eben,
Sorry for misspelling your name. Yes, I’m happy to talk to you or any typophiler offline. My address is kevlar at microsoft.
Cheers, Kevin