blackletter legibility

[Guest]

Discussion split from here.

 

Hey, what do you think about letter "k" and "t" in this font. It's so little difference between them. It's unbelievable that people really used it in newspaper. "S" and "V" glyph are intresting too. It's really unreadable for someone who never saw Faktur before. But it's good for me. I think it will make the stronger effect in my project. You know... If it looks more "exotic" it will cause bigger culture shock :D Well, there is a lot of technical problems. For example - paper. I want to use thick offset paper, but it's to white. I don't want to use colored paper because it won't look realistic. So I need to print it early enough, put paper in sunny place and... wait. I think that would work but... well, you never know, right? Second problem - photos. I can't just convert it into grayscale. The raster will be to good. I need to simulate technology from '30 so there is a lot of work waiting for me. But there is one thing good for me - I'm working in a printing house, so I've got a lot of paper and professional RIP there. There's a lot of theoretical work too. Well, last time I'm wondering, in what type of culture we are participating in with present forms of communication design? What do you think? Oh... and of course, I'm attaching an image with "k", "t","S" an "V" :D

[Ralf Herrmann]

On 15 March 2015 at 10:59 PM, Ponczi said:

Hey, what do you think about letter "k" and "t" in this font. It's so little difference between them. It's unbelievable that people really used it in newspaper. 

Believe it or not, but there are still people today that claim such blackletter fonts offer superior legibility over Roman typefaces. 

[Wrzlprmft]

Also, you do not need to weather the paper before printing on it. We keep some old unused paper around just for such purposes.

 

As for the k, t, S and V and other difficult to read blackletter letters: That’s why I created these (cv01–cv10 or ss01 of UnifrakturMaguntia):

[Wrzlprmft]

On 15 March 2015 at 11:07 PM, Ralf Herrmann said:

Believe it or not, but there are still people today that claim such blackletter fonts offer superior legibility over Roman typefaces. 

Well, it’s not that much more absurd than claiming that roman typefaces are better readable than blackletter ones. What is really weird is that these people think it’s self-evident or that they can actually prove it.

[Kathrinvdm]

Isn't it all a matter of being exposed or not exposed to something? If I was used to read blackletter all my life and suddenly come across some text written in roman letters - for sure it would be a miracle for me how to read it. And the other way around: same thing. But with some exercise one manages to read blackletter quite fluently, from what I have experienced.

What would be the right way to scientifically proof the superiority in readability of one of the two letter systems? I have no clue, but I think it would be fairly difficult to test, because most people nowadays are so used to read roman letters. Maybe it works if one tests people that come from a completely different letter system background, such as Chinese or Cyrillic? *shrug*

Edited March 17, 2015 by Kathrinvdm

[Þorsten]

Perhaps not if you look at all of blackletter, since there are too many variations. But if you were to look at just Fraktur vs. typical examples of Roman type, Fraktur has a fundamental problem Roman type simply doesn’t have: those awfully similar glyph shapes for different letters.

There is plenty of evidence that tells us that lack of distinctiveness between symbols (incl. letters) that carry different meanings is a problem: for people with all kinds of cognitive impairments (incl. but not limited to dyslexia, reduced vision), children who learn the alphabet (I can see this in Þorstelinchen every day) as well as people in general under less than ideal conditions (reading a highway sign through a rain-streaked windshield, reading a badly degraded old newspaper etc. etc. etc.)

When we argue about ẞ (capital ß), we frequently reject designs for being “too similar” to B. Nobody even questions why this might be a problem; it’s just assumed as a given. I would argue that even the worst B/ẞ pairs we’ve seen probably are still a lot more dissimilar than J/I and ſ/f are in a typical Fraktur.

The degree of dissimilarity between letters can be measured objectively, of course, using image processing algorithms. Once we accept the premise that lack of distinctiveness is a problem, I’m not sure what else there is to do.

[Ralf Herrmann]

On 17 March 2015 at 1:07 AM, Þorsten said:

Perhaps not if you look at all of blackletter, since there are too many variations. But if you were to look at just Fraktur vs. typical examples of Roman type, Fraktur has a fundamental problem Roman type simply doesn’t have: those awfully similar glyph shapes for different letters.

By the way: that difference between I and J is a rather recent invention. The typical Fraktur, as used over centuries, would use the same glyph for I and J. 

[Wrzlprmft]

Fraktur has a fundamental problem Roman type simply doesn’t have: those awfully similar glyph shapes for different letters.

Roman type is not free from that either. There is, most prominently, I and l (which look identical in this edit interface). Some dyslexics have trouble telling b, d, q and p apart – a problem that is probably alleviated in blackletter. You could also start arguing that blackletter has more ascenders. I am not saying that you can draw a direct comparison between either of these aspects, or am even claiming that one is actually better; but it’s just not that simple.

The only way to decide this question is to have hundreds of subjects only exposed to one type of fonts from birth or strictly equally exposed to both types of fonts and then make reading experiments with them at a certain age. As this is highly unethical, I am being agnostic about this question: We cannot know which is better.

Also remember, that there is still no solid evidence as to whether serif or sans-serif is better readable for long texts.

Edited March 18, 2015 by Wrzlprmft

[Þorsten]

Well, I would submit that none of these — very real — issues (with, e.g., I/l and b/d/p/q) rise to the level of difficulty presented by classical Fraktur typefaces, for several reasons:

1. In many “classic” Roman (serifed) styles, I and l differ in three, four or more aspects: e.g., height, slope of the top, top left serif, top right serif (and sometimes even stroke width). In traditional Fraktur typefaces, the most critical letter pairs (I/J and f/ſ) differ in one aspect at most, which may be incredibly difficult to spot at body text size (the right part of the f-bar is but a tiny dot at that size.)
   
2. So what about grotesques and other sans-serifs where I and l are identical? True, this is a legibility concern. However, as we all know, it is entirely possible to design typefaces where all letters are clearly distinguishable by multiple aspects — without deviating from what is considered canonical for the style. Can you say the same for Fraktur? In any case, I/l-confusion is less likely in real-world texts anyway due to context. (An uppercase vowel and a lowercase consonant are much less likely to be mistaken for each other via-a-vis two lowercase consonants.)
3. I haven’t even addressed uppercase Fraktur letters. Doesn’t the widespread custom of setting all-caps acronyms etc. in Roman type pretty much amount to an admission that Fraktur caps (that infamous assortment of “elephant trunks”) need context to be legible?

So could blackletter typefaces be made (much) more legible? Probably yes, but if anyone tried, those changes sure haven’t gained any degree of meaningful acceptance.

[Ralf Herrmann]

 

Also remember, that there is still no solid evidence as to whether serif or sans-serif is better readable for long texts.

​But that is very different question. The difference between sans and sans-serif is mostly a matter of style. What matters most for legibility is the structure of the letter. The skeleton and it’s proportions one could draw with a stick in the sand. (And secondary features like weight and spacing which might affect if the skeleton is clearly visible.)

In that regard there are clear differences between blackletter and Roman typefaces, but not necessarily between sans and sans serif. The legibility of a character does not depend on the fact if a stem ends with little feet or not.  

[Wrzlprmft]

​But that is very different question. The difference between sans and sans-serif is mostly a matter of style. […] The legibility of a character does not depend on the fact if a stem ends with little feet or not.  

And yet, you very often read claims that serif or sans-serif is easier to read in certain situations, e.g., I am pretty certain that that Wayfinding Sans is a sans for a good reason. Most importantly, you very often hear that serifs are better for long, printed texts, as the serifs guide the eyes (or similar). However, if I am not mistaken, there is no really compelling argument, let alone a scientific study to back up this claim. Sure, the question is different, but it’s similar enough to illustrate how little we actually know or can decide regarding readability.

[Ralf Herrmann]

​

And yet, you very often read claims that serif or sans-serif is easier to read in certain situations, e.g., I am pretty certain that that Wayfinding Sans is a sans for a good reason. Most importantly, you very often hear that serifs are better for long, printed texts, as the serifs guide the eyes (or similar). 

​To me, these things are a more a sign of how poorly we define and use our terms to describe such issues. Reading a city name of 6 characters on a road sign 300 meters away or reading a 300-page book — those are very different things. If we just classify fonts as “more easy to read” and “less easy to read” and then apply it to such real world situations, it will cause confusions. But that in turn doesn’t mean we can’t know anything unless there is scientific proof. 

[Wrzlprmft]

But that in turn doesn’t mean we can’t know anything unless there is scientific proof. 

But there should at least be an argument based on something that is scientifically backed up. Personal experience and gut feeling are just too unreliable.

[Þorsten]

On 21 March 2015 at 8:57 AM, Ralf Herrmann said:

The legibility of a character does not depend on the fact if a stem ends with little feet or not.  

Here too context as well as the literacy of the reader probably matter. Text read by competent readers certainly is a completely different animal, but here is a single data point of a very different context …

For what it’s worth, my 2-year-old appears to have a slightly easier time regognizing certain individual characters when they come with prominent serifs. The backstory: when she was 20 months old, I noticed — much to my surprise — that she recognized all capital English letters. She most likely learned them from occasionally watching sign-language videos (geared towards children) that we got at the library. At the end of most videos, the presenter would show the signs of the manual alphabet¹ while the corresponding letters were displaying on the screen (in white Comic Sans MS on chalkboard green.) A short while later, she discovered a simple game on a library computer, which simply displayed any typed letter on the screen. When I built a simple web page doing something very similar², I experimented with various typefaces, incl. Comic Sans (which she already knew :-(), designated for-beginning-readers typefaces like Andika, various “soft” Comic Sans alternatives such as Overlock, straightforward sans serifs and others. The typeface where she hesitated least when being shown characters that I had typed turned out to be Canapé Serif. (Canapé has a two-story a and a three-story g, by the way, which didn’t appear to confuse her in the slightest.)

 

download at MyFonts

Andika: http://scripts.sil.org/cms/scripts/page.php?site_id=nrsi&id=andika
 

 

On 22 March 2015 at 7:55 AM, Wrzlprmft said:

But there should at least be an argument based on something that is scientifically backed up. Personal experience and gut feeling are just too unreliable.

Maybe so. I still take empirical observations over post-hoc rationalizations of ideological preconceptions.

 

1. http://en.wikipedia.org/wiki/American_manual_alphabet

2. http://type.dt1.org/abc/

(For some reason, the insert-link function of the editor does not work for me.)

[Ralf Herrmann]

For what it’s worth, my 2-year-old appears to have a slightly easier time regognizing certain individual characters when they come with prominent serifs.

​Thanks for sharing that story. That’s a fascinating observation worth exploring in the future!

[Wrzlprmft]

​I still take empirical observations over post-hoc rationalizations of ideological preconceptions.

Science is nothing but empirical observations and deduction from theories founded on empirical observations. The problem is just that it’s at times bloody difficult not to cloud one’s obervations with wishful thinking or fall victim to self-fulfilling prophecies – which is exactly what the ideologists we are talking about here are doing.

[Hans]

​Believe it or not, but there are still people today that claim such blackletter fonts offer superior legibility over Roman typefaces.

​They are, at least for German, because German words are usually longer and so more Fraktur letters can be seen at one glance. In addition Fraktur uses letters like "langes s", which make certain German words easier to understand.

[ÖNordling]

Black letter fonts offer a decorative feeling in newspaper mastheads/logotypes. Legibility might not be the first aspect. Some swedish samples here from a speech performed by me and Fredrik Andersson at SNDS in Stockholm 2007. 

Barom1.tif

BoråsT1.tif

NT1.tif

[Ralf Herrmann]

Black letter fonts offer a decorative feeling in newspaper mastheads/logotypes.​

Certainly true today, but blackletter fonts were indeed used for regular book & newspaper typography for centuries.

I agree however, that many design aspects of blackletter have their roots in decoration not in a search for maximum legibility.  

[Chris Lozos]

The other thing to remember is the degree of literacy back when Blackletter was dominant.  There were far fewer readers in the world and far, far fewer things to read.  There was not the speed of life that we now experience or the need for very quick reading.  Most people got information from the spoken word.  For us to compare readability from centuries past is not possible.  Today, however, we do not need to read all things equally quickly.  A highway sign must be quick reading, but a movie poster does not. We have the luxury of dwelling on some things we like and even ignoring others [junk mail].

[Hrant Papazian]

And oldie but –hopefully– goodie: http://www.themicrofoundry.com/ss_fraktur1.html

I believe blackletter can still inform the functionality of Roman type, if mostly tangentially.

[Ralf Herrmann]

Just noticed: the word hyphen doesn’t really go well with the idea of superior blackletter distinguishability/legibility. 

[Hrant Papazian]

Yes, the classic blackletter "y" is no good for text, but the descending "h" is golden.

[Wrzlprmft]

Just noticed: the word hyphen doesn’t really go well with the idea of superior blackletter distinguishability/legibility.

Well the claim is usually only made with respect to the German language and thus unaffected by your finding. (Not that I support the claim.)

By the way: You seem to be working on something blackletter-related. May I ask what?