Least legible/readable typefaces (Print)

Mat Lucas's picture

Fellow T’philes,

In my quest for the answer to this question I have come across countless rants and fanboy’isms behind the “best” typeface, but what are your opinions or solid gold facts about the worst readable faces? I’m thinking mainly in print for copy, not so much titling.

What are the worst offenders?

Thanks,

Mat.

Nick Shinn's picture

IIRC it was Degas who said "Give me a palette of various hues of mud, and I will paint you a rainbow".
In other words, a good typographer can make any typeface read well.

However, just as some musical instruments are harder to master than others, some types are more intractable.

Mat Lucas's picture

That is a great quote! I'm a believer in “there is no bad type, only bad usage of type.” From purely a legibility view point though, e.g. counters too large or a large x-height.
Are there any type faces that are notorious for their poor legibly?

Ha i'm getting a feeling of that Carson spread for Brian Ferry, thanks Claus.

This will probably end ironically, being a student I have more than likely used ones that might be mentioned more than once!

Mat.

Edit: Beat me to it Claus!

Nick Shinn's picture

"I'm not bad, I was just drawn that way" -- Jessica Rabbit

Didones were born with a bad reputation, following the famous Ben Franklin incident (which involved Baskerville, a transitional).
They are indeed hard for the type designer, typographer and printer to get right.
That's one reason I did a Scotch Modern revival, to try and do a version that would make 21st century text in the style look more like the better examples from the mid 19th century.

Mat Lucas's picture

Nick, I can definitely see the problems with the legibility with the Didones , thanks for the heads up.

With your “Scotch Modern” how much did legibility come into question when designing it?
I feel it has more legibility than Didot, for example, if only from a screen preview of it.

Mat.

Nick Shinn's picture

My Scotch Modern is designed to create a facsimile of typesetting in a particular 1869 book, which I thought was highly legible/readable, so I didn't concern myself with legibility per se. However, I did add some kerning, which is expected these days, and necessary, given the sophistication of h&j software.

butterick's picture

Perhaps there are no bad typefaces, only insufficiently skilled typographers, but I've never been glad to see Melior.

oldnick's picture

Perhaps there are no bad typefaces, only insufficiently skilled typographers

Again, I return to my tool analogy: that works best which is suited to its purpose. Many a typeface which shines in print may suck on the screen, and vice versa. The worst offenders are those errant knaves who know not the difference...

Adam Shubinsky's picture

I just finished reading a book set in Utopia (with tight leading!), and I can assure you that it was a rather dystopian experience (which was a shame since it was a rather good book, content wise)

I would rather drink Cyanide and wash my eyes with Nitric-Acid, than read another book set in that inelegant monstrosity.

10/10 for Legibility
0/10 for Readability

Nick Shinn's picture

Did you ever read The Globe and Mail when it was set in Utopia (2001-2007)?

charles ellertson's picture

Perhaps there are no bad typefaces, only insufficiently skilled typographers, but I've never been glad to see Melior.

P. J. Conkwright used a fair bit of Melior . . .

Adam Shubinsky's picture

- Did you ever read The Globe and Mail when it was set in Utopia (2001-2007)?

Nick, if your point is that there are no bad fonts, just misused ones, then I get it (and for the most part, agree). I have no doubt that Utopia has its uses, long lines of extended text just aren't one of them. The thing is that since it's makers have, on more than one occasion, championed its book setting qualities, I have no qualms about denouncing it as a "bad" font.

As for reading the Globe and Mail, I only started reading it in May 2007, so I guess you "deprived" me of my Utopian pleasures.

William Berkson's picture

I do think that a lot of fonts are unsuited for extended text at small size, which is what I think you mean by "copy." And the market agrees, as there is a much smaller range of fonts used for books and magazines for extended text compared to the fonts used for display and short copy.

How often have you seen extended text in Curlz? I would nominate this as unsuitable for extended text. I would challenge anybody, including Nick Shinn, to set a book page successfully in Curlz, with the text block in extended text, small size, many lines.

Nick Shinn's picture

What do you mean, "text block in extended text, small size, many lines" ?
The stipulation is "copy" (as opposed to titling), there is no mention of handcuffing the typographer to a tombstone.

William Berkson's picture

Most magazines, and almost all books have such text. I am just wanting to see Curlz work in a situation that the majority of text is set in. No handcuffs, just a realistic demand. Show us.

Adam Shubinsky's picture

In truth, I believe that there are probably no more than fifty typefaces that are truly suitable for setting books and magazines. Add in a range of diacritics, and suitability to modern printing technologies, and that list gets whittled down to around half that number (William, I believe that your Caslon would definitely be on this shortlist).

It never ceases to amaze me how few fonts there are that are easy to read (in book settings), possess an elegant and inviting form (which in my eyes means not too bland and not too fancy), have the correct proportions, are endowed with good metrics, include all the necessary feature sets, and look good when printed offset (or even laser). Yet, even though this list is rather small, some book designers still manage to screw things up!

butterick's picture

On reflection, I have to reject the premise that there are no bad typefaces. And by "bad" I don't mean just goofy or ugly or badly drawn. I mean typefaces that are fully considered and skillfully rendered but that just don't do their job well.

To say that there are no bad typefaces is to imply that type designers cannot fail: no matter how badly they screw up, a page designer can still redeem the work. But type designers, like all designers, are fallible.

But I also don't believe that one can paint a rainbow with mud.

Nick Shinn's picture

Most magazines, and almost all books have such text. I am just wanting to see Curlz work in a situation that the majority of text is set in.

Bill, you have an idée fixe as to the nature of "text type".
I will admit that most text is set in the codex slab, on account of books being so long, but most pages of most documents containing text type are not like that.

To say that there are no bad typefaces is to imply that type designers cannot fail

That's setting the bar rather high. What about the majority of typefaces which are average or mediocre?

It never ceases to amaze me ... in my eyes...

Well yes, wouldn't most discriminating typophiles have a taste that narrows the field to their personal favorites?
But wouldn't there be some variance of preference amongst them?
I don't doubt that there would be a top few dominating, but according to Zipf's Law, the top fifty of a group of typophiles would have many hundreds of faces.

William Berkson's picture

>idée fixe

Instead of irrelevant efforts at mind-reading, why don't you show us a typical book column of say 11/14, or magazine column, say 9/11, set in Curlz, and make it as successful as typical widely used text types such as Minion and Sabon, set next to it?

Nick Shinn's picture

Bill, I'm playing by the rules of this thread: Mat said, "I’m thinking mainly in print for copy, not so much titling."

Why do you assume that "copy" means a solid page of text, rather than, say, a paragraph in a deck or advertisement?

Remember a recent thread where we discussed Paul Rand?
That was where you stated that the 8 pt. type of his business card was display.

I'm not reading your mind, but going by the evidence of your posts to Typophile, where when the subject of "readability" crops up, you always assume it refers to extended tracts of book or magazine text, e.g:

... why don't you show us a typical book column of say 11/14, or magazine column, say 9/11, set in Curlz, and make it as successful as typical widely used text types such as Minion and Sabon, set next to it?

Again, I have to ask, is that the only spec there is that can assess the readability of text type?
Remember Paul Rand's book set in Univers 75?

I'm not suggesting that a typographer can turn any type into a workhorse text face, all I said was that a good typographer can make any face read well. Part of that is setting it with sensitivity to context and content, not just dumping it into a generic spec.

William Berkson's picture

Nick, I conceded in that thread that there is ambiguity about the meaning of "text" as it usually involves both small size and extended text, but you can have one without the other. And I maintained, I think correctly, that may fonts, including Univers, are not optimal in extended text at small size. And I would add that some, including Curlz, are just horrible.

Mat here contrasted "print for copy" and "titling." "Copy" usually means extended text, but Mat can clarify that for us.

In your initial post, in response to a question about "copy" you wrote: "a good typographer can make any typeface read well." Given the context of "copy", that meant "read well in copy". Now you back off and say just that a typographer can find some conceivable situation in which it can read well. But that was never at issue, because the question wasn't whether a few letters or words can be made readable, but whether there are faces that are bad for "copy", extended text.

Changing the meaning of the question to win an argument is known as the "fallacy of equivocation". In fact, by saying "I'm not suggesting that a typographer can turn any type into a workhorse text face," I think you have conceded that Curlz is not good for "copy"—thus agreeing with my point while not admitting it.

Adam Shubinsky's picture

Well yes, wouldn't most discriminating typophiles have a taste that narrows the field to their personal favorites?

Nick, you are missing my point, I was referring to something that specifically transcends the notion of "personal favourites". The whole point is that typographers personal favourites notwithstanding, there are ultimately only a small number of fonts that are objectively best suited for book setting, and possess the myriad of qualities and features that make reading any book set in them, a smooth and pleasurable experience.
If a typographer tries to force a certain typeface on a book, for any reasons other than readability (e.g. "This font is so cool", or "This high-contrast font comes from the same historical period as the plot of the book, so I should use it as a display of typographic savviness"), the end result would most likely be taxing on the reader.

...according to Zipf's Law, the top fifty of a group of typophiles would have many hundreds of faces.

I do not think that Zipf's law can or should be applied to this particular context.

BTW, I tried setting a book page in Curlz (11/14) and, as predicted, it was a nauseating experience...

Mat Lucas's picture

“Mat here contrasted "print for copy" and "titling." "Copy" usually means extended text, but Mat can clarify that for us.”

Bill, yeah sorry about the ambiguity it was not intended, I was thinking of typefaces that would be poor for say extended copy use in a book or even ones that would struggle in the context of a magazine as a piece of body.

Sorry for the late reply loads on at the moment, great dialogue to read through this morning over coffee though!
I have to agree (not that my opinion should be validated as something even remotely informed) that a copy in “Curlz” is a little disheartening at best. Also though I have never seen this set for copy in the wild, maybe titles for amateur dance groups etc. on a nice pink low gsm stock.

It is repulsive, and I’m sure not designed for this purpose (or many others) but it does re-enforce the idea of “no bad typefaces, only bad usage” as anyone who would set this for copy would be insane, other than for the amusement of this thread!

Thanks Riccardo, I have never seen those faces before.

Thanks also for Nick recalling a book set in "Univers" it's something I will look into.

Dan B.'s picture

I just finished a book of about 120 pages set in [wait for it] Lubalin Graph. Painful experience. The only thing that kept me going was that it was required reading for a class.

dtw's picture

Riccardo, you beat me to that one. :^)

Nick Shinn's picture

@Bill: Changing the meaning of the question to win an argument is known as the "fallacy of equivocation".

Mat said "in print for copy, not so much titling."
I didn't change the question, I interpreted it quite literally, with "copy" meaning any text that is not titling.
In typography, there are many genres of text, such as decks, sidebars, captions, pull-quotes, &c., not just book text.
Narrowing one's interpretation of the question to suit one's prejudices is known as "Reductionism", if not "Straw Man".

...the question wasn't whether a few letters or words can be made readable, but whether there are faces that are bad for "copy", extended text.

That was your assumption, which proved to be correct. Of course I suspected that was what he meant, but decided to take the literal side of his ambiguity, because I'm fed up with Readability Reductionism, and the way that its adherents diss the readability of any face that won't set a book, also dissing the typographer's contribution to readability.

@Adam: I was referring to something that specifically transcends the notion of "personal favourites".

No you weren't. You said "...in my eyes"! There is no object measure of readability for text faces.

I do not think that Zipf's law can or should be applied to this particular context.

Why not? It is very good at cultural listism.
If you polled typographers on the "top ten book faces" you would get a Zipf curve, with Caslon and Bembo at the top and many types, such as Electra and Dolly, appearing once or twice.
And if you went to a library, Times would be way ahead.

dezcom's picture

Anytime someone tries to establish an absolute for things that are somewhat fuzzy in nature, they are asking for an argument. There are numerous reasons book publishers choose a typeface for book text. One of them simply being "workflow". They have used a face or faces so many times that page-out is very predictable. Therefore they see no reason to switch since they don't get a dime more profit for their trouble. Yes, Curlz is a lousy book face but their may be occasions when an author is more interested in establishing a character than having what they say be easier for scientists to agree upon for readability of content. This kind of thing is more likely in the case of advertising than setting a typical pulp-fiction novel. If someone wanted to set a scenario where the speaking character was a shallow, ditzy teenage girl and had little of value to say, Curlz might fit the bill perfectly.
We all know the tenuous history of readability studies and how much disagreement there is about cause and effect--correlation is not always causal. I guess I just don't see much value in beating that horse yet again. No real usable information comes out of the argument. This thing always surfaces out of "best "worst" "most" "least" "never" "always" kinds of lists. e. e. cummings only used lowercase and Massimo loves Helvetica and Bodoni. Yup, so graphic designers, editors, publishers, advertising people, and scrapbookers will continue to pick typefaces either well or badly as they always have and any pronouncement of an absolute will continue to be either ignored or argued about without changing a thing.

William Berkson's picture

>I'm fed up with Readability Reductionism, and the way that its adherents diss the readability of any face that won't set a book, also dissing the typographer's contribution to readability.

Ok, I think your argument was bad, but I see you do have a valid point you were trying to make. Even if a font is not ideally suited as a book face, it can have better readability than many competitors for its purpose. For example it could be a "text" sans that works better as a contrasting type for short-measure call out boxes in a magazine or news paper. Or it could be a companion italic. These don't have to be ideal for extended text, but still it is an achievement to make them *relatively* readable for their function. Fair enough.

However, I was never arguing against that being true.

metalfoot's picture

I once read an entire book--- about 136 pages, if I remember correctly, set in tightly kerned and leaded Clarendon. My eyes are still recovering from that a year later.

Nick Shinn's picture

...it can have better readability than many competitors for its purpose.

Right. Readability is document-specific. And for non-book text typography, "book" types can have poor readability, because the reader may take one look at a text block and, not being in the mood for serious endeavor, decide against immersion without reading a word. That too is a readability experience.

Adam Shubinsky's picture

Nick, by invoking Zipf's Law you are in fact proving my point (albeit in a roundabout way). If the distribution of book-faces preferred by typographers follows a Zipf curve, and you end up with typefaces such as Bembo, Caslon, and Garamond occupying the top of the curve, then perhaps there is some uniquely inherent quality (or set of qualities) to these particular typefaces that makes them universally appealing in the context of book reading.

If you ask these typographers to name their top 20 typefaces for books, you will probably end up with three to five top choices with a myriad of other choices occupying a widening shallow distribution. You seems to look at this as an indication of the variability of preference, and this variability would in turn appear to support the notion that such choices are ultimately highly subjective. Yet, the information could be interpreted the other way round—that there are only a small number of typefaces that possess a universally agreed upon quality. This in effect was the essence of my core argument.

It stands to reason (and statistics), that if we have some superb all-round book faces, and others that are just ok, then there must be some (Curlz is obviously an extreme example) that are bloody awful. The question that started this thread was simple: Can we name some of the font that are least suited to these types of settings.

One last thing though; the claim that certain typefaces (e.g. Curlz) could, under certain extremely rare circumstances, be called for in book text, and therefore could not be considered to be inherently bad for such setting (the strong version of the "it all depends on the context" argument), is just a cop out. After all such an argument could be extended to all types of symbolic forms including dingbats.

Chris Dean's picture

@ Mat Lucas: Is this for personal, professional or academic purposes?

By “solid gold facts” do you mean scientific studies and empirical data? That’s a tricky question. There have been many scientific studies comparing fonts to determine their influence on human performance but none spring to mind where someone was trying to find the worst. If I were to answer the question of “worst offenders” I would start by reading as many studies as I could find comparing fonts, look at all the losers and try to find common qualities. From there I would design an experiment, which would inevitably lead to a series of experiments.

If you already haven't you may wish to read this thread: BBC News - Making things hard to read 'can boost learning'

JoergGustafs's picture

So what did we learn so far? Lubalin Graph sucks as a book face and Utopia is just as much of a ‘worst readable (…) worst offender’ as Curlz, at least when used ‘mainly in print for copy, not so much titling’.

;)

Nick Shinn's picture

We did?
Utopia is a much respected text face for newspapers.
Oh sorry, that doesn't count, "copy" mean book text.

**

Adam, the Zipf curve describes the herd instinct: some things become very popular, quite apart from any quality of merit, because that's the way culture is.

For typefaces, that means some faces will become the go-to faces preferred by typographers.
How much does this have to do with the inherent readability of types for the end user, who has no say in the matter?
Not a lot.

William Berkson's picture

Utopia can look fine in newspapers. The example discussed must have really had a bad typographer to mess it up. The best text face can easily be destroyed by misusing it. The look is always the face + the design. They both need to be good. If only one is, you've got big trouble.

kentlew's picture

> The best text face can easily be destroyed by misusing it.

And by using Adobe’s Optical Metrics. I’ve seen plenty a decent face thrashed by the this setting. This has become disturbingly common in magazines. And I’ve seen it in a few books now, too.

hello seb's picture

all i can say is that i'm no fan of Hobo Std

JoergGustafs's picture

Nick Shinn:

Actually I don’t share Adam Shubinsky’s cordial antipathy towards Utopia, to me it’s a great typeface even for longer texts.

I was trying to express my disagreeance with the (IMHO) rather pointless nature of 1.) the quest for the ‘worst offender’ without really restricting the parameters and 2.) the practice of throwing everything in the same ‘worst readable’ pot, Utopia together with Curlz, Hobo and what else.

Adam Shubinsky's picture

- Adam, the Zipf curve describes the herd instinct: some things become very popular, quite apart from any quality of merit, because that's the way culture is.

Nick, I don't mean to be picky but... Zipf curves are not a description of herd instincts per se. They describe a common phenomena that appears in a variety of setting in the natural world—just like its close relative Pareto's Law (the so called 80-20" law). Herd instinct and behaviour on the other hand is an emergent property of the basic rules that govern the behaviour and interaction of individual herd members.

- some things become very popular, quite apart from any quality of merit, because that's the way culture is.

I agree! A brief review of the popular programs displayed on television networks or the latest (mostly 3D nowadays) movies, certainly proves that point. Another thing that seems to be rather popular today, is bashing and trashing the old classic fonts. It would seem at times, as if the typographic herd instinct collectively frowns upon those refuseniks who still cling to those Caslons, and Bembos, instead of embracing the brave new fonts of our inventive times.

Just to clarify a point, I do not hate Utopia (I don't even "hate" Comic Sans). I hate that it is used in book settings (and often with tight leading to top!), it is an eye sore, just like the man who attends a white tie affair, dressed in shabby jeans and rock climbing boots. If you ask me who is the worst dressed person in the room, then I would be inclined to say that it is the man with the shabby jeans. That is not to say that I have laid a critique against the use of jeans in general!

dezcom's picture

and here I sit reading this wearing my shabby jeans as usual. Maybe that is why I never get invited to white tie affairs--that and the fact that I don't know a soul who has either been to one let alone hosted one.

dinazina's picture

I'm reading a paperback mystery that is set in Galliard. The Roman is very pleasant, but the italic (which is used often) is so different and extreme it is jarring to the eye. I especially notice the strange Y, as in "You fool."

@kentlew: Can you please explain why Adobe's "optical" setting for text is bad? It appears to tighten the kerning of some letter pairs, which generally looks like an improvement to me. But what do I know.

dezcom's picture

What Adobe calls "optical" I call hit-or-miss. Sometimes the Adobe thing looks reasonable (mostly at large display size) but sometimes you just say WTF were they thinking. Perhaps I am of little faith when it comes to any automatic cure-all but I find it more trouble than it is worth. I trust my eyes more than their formulas.

William Berkson's picture

Sometimes the Optical spacing can be of help, for example if swashes are not kerned, or with "camel" casing like MicroSoft. (Lower case—uppercase kerning is unusual.) But even there it is better to hand kern if the typesetter has learned to do it.

But aside from these kinds of exceptional cases, it is a menace, as Kent has shown with examples in other threads, with the capacity to wreck a well kerned font. The worst thing is that the idea has been propagated that this is a "superior" kerning, which is as a rule the opposite of the case.

It would be best if Adobe would either cut it altogether, or at least explain it properly.

Chris Dean's picture

Comparing the two different settings using measures of speed and comprehension would make a wonderful study (and probably save a lot of time for software developers and typographers alike).

kentlew's picture

Dina — The bottom line is that Adobe’s Optical metrics ignores the metrics built into a font and completely overrides and subverts the spacing that a type designer has spent time designing. For a well-considered text face, this is not a good idea.

At best, there is little noticeable difference. At worst, it can completely disrupt the rhythm of a face. I am perhaps a bit touchy about this topic because Whitman happens to be one of those that suffers horribly at the hands of so-called “optical” metrics. I do not proclaim that my fitting of Whitman is flawless or perfect. But what Adobe’s algorithm chooses to do with it is a travesty.

A more in-depth critique of Adobe’s Optical metrics may be found embedded in this thread:
http://typophile.com/node/54310#comment-327297

type addict's picture

Not Caslon by Mark Andresen. NOT legible.

I wouldn't recommend using this typeface for
road signs!

butterick's picture

Trump Medieval — another one that gives me fits

Té Rowan's picture

IMAO: The grunge faces.

Syndicate content Syndicate content