Can you believe a woman got fired from her job for using all caps in an email? There’s something about all caps text that turns people off. Using it in a social context means you’re yelling. But using it on your website means bad readability for your users.
Many websites use text in all caps to emphasize their message. However, what they’re actually doing is de-emphasizing their message because text in all caps reduces the shape contrast for each word.
The shape of any word in all caps for any type of font, sans-serif or serif, is a rectangle. This means that text in all caps show a parallel edge at the top and bottom of a word, giving it low shape contrast. But text in title style capitalization show multiple adjacent edges at the top and bottom, giving the words high shape contrast.
The more nonparallel edges your text has, the higher the shape contrast it has. High shape contrast makes words easier for users to recognize. If you want to make your text easier to read, consider using title style capitalization instead of all caps on your website.
When is it okay to use all caps? All caps are fine in contexts that don’t involve much reading, such as logos, headings, acronyms, and abbreviations. But when your message involves reading, don’t force users to read words with bad shape contrast. The caps lock key is a key that designers should rarely use.
In emails, using all caps is a sign of bad manners. In design, using all caps is a sign of bad readability. Know when and when not to use all caps, and you’ll have a better chance of keeping your users and your job.
The major missing element in this analysis is learned behavior. How much does the long dominant use of mixed case in the English language affect readability?
In Russian it is often the case, perhaps dominantly, that all caps mixed large/small (first letter of sentence, names, etc. are large caps followed by smaller caps) is used. Mixed caps does include a distinguishing sized initial letter, but it is then followed by all square letters – effectively falling somewhere between English all caps and English initial-caps.
Does that mean Russians are unnecessarily hindering their capability to read – or does it mean that long term dominant behavior plays a significant role in readability?
The major missing element in this analysis is learned behavior. How much does the long dominant use of mixed case in the English language affect readability?
In Russian it is often the case, perhaps dominantly, that all caps mixed large/small (first letter of sentence, names, etc. are large caps followed by smaller caps) is used. Mixed caps does include a distinguishing sized initial letter, but it is then followed by all square letters – effectively falling somewhere between English all caps and English mixed case.
Does that mean Russians are unnecessarily hindering their capability to read – or does it mean that long term dominant behavior plays a significant role in readability?
I’m reminded of the interesting behavior where readability is surprisingly maintained (though clearly affected) when only the first and last letters of a word are consistently positioned.
http://blogs.msdn.com/b/fontblog/archive/2006/05/09/594050.aspx
Your facts are wrong, or you meant some other language I am unaware of, but there’s no such thing as all caps in russian. In fact, in russian it is grammatically correct to capitalize only the First letter of the title, not Every Word As In English.
This is why in the UK, all placenames on roadsigns are printed capitalized, not uppercase – it’s easier to recognise the names from the shape they make than from actually reading them if travelling at speed.
This was all discovered back ages ago. it’s a surprise how these things come back around
See Word Recognition for a very good paper that debunks the word shape model, which has, apparently, long since been superseded by the parallel letter recognition model.
The word shape model is still talked about a lot, but Daniel is right, it’s out of date. I wrote a blog post that summarizes the research on this topic, called It’s a Myth That All Capital Letters Are Inherently Harder to Read: http://bit.ly/8QUScv
The title of the article you link to is misleading to people. It’s not a myth that all capital letters are harder to read. In the conclusion you say, “All capital (uppercase) letters are slower for people to read, but only because they aren’t used to them.” This confirms that it’s not a myth.
It’s hard for people to read because of the word shape or “parallel letter recognition” if you want to get semantically technical. The word shape contributes to letter recognition, so the shape of the word is still a factor. The insight that you are trying to offer is that people recognize words by anticipating the sequence of letters, which is why having letters or words that vary in its shape is important.
I welcome other point of views that add value to discussion, but please be more careful with your headlines, as to not mislead people. And if there is something that is different than what is mentioned highlight it so people can see. It helps.
I think you miss the point and one important word in the headline (viz., “inherently”). Yes, “All capital (uppercase) letters are slower for people to read, but only because they aren’t used to them.” But that does NOT mean that they’re *inherently* harder/slower to read. I guess that most people are not used to texts set in fraktur, so that would be rather hard to read, but how hard would it be for someone brought up on such texts?
(And the Hebrew alphabet lacks a case distinction and could be considered uppercase only.)
Furthermore, what is the point of promoting an out-of-date, shown to be wrong, theory. Is the rest of you’re writings also based on out-of-date stuff?
Interesting that the author of the blog post supposedly debunking word shape as a readability issue…publishes the blog in lower case instead of in ALL CAPS. I wonder why?
In the days of teletypes in radio and TV newsrooms, the information came across in ALL CAPS and was read by announcers fairly easily – perhaps they were used to it?
A n d . . . w h a t o f k e r n i n g ?
I DON’T GET IT.
What about no title casing? That was pretty hip for a while, and lots of sites still do it. So like buttons would just be “save” instead of “Save” or “SAVE”. Or I remember in my tech writing classes being taught that initial caps in headings slowed down readability too.
Metro UI on Windows Phone actually uses lowercase-only for titles quite often, e.g:
http://windowstotal.com/wp-content/uploads/2011/03/ebay-mobile-windows-phone-7.png
Totally agree! Great analysis!
The apostrophe key should also be used with caution 😉
I would have put an apostrophe after “used.” 😉
Cheers, and thanks for the well illustrated explanation
Now begs the question of whether words in all caps with the first letter of each word bigger than the rest is easier to read or as easy to read as not using all caps.
As a dutchman I was tought to read phonetically because unlike English many other langauages are phonetic and have consistent spelling. I have noticed over the years that English speakers are not taught to read phonetically and appear to have great difficulty reading foreign words or in fact any unfamiliar words. I do not have this problem as my brain reads words like a string of beads, I read each letter seperately and then form a word; English speakers appear to me (I wonder is anyone has done research on this) to read by recognizing entire words and hence have problems with recognizing new words they have not learned yet and I wonder if this may be the reason UPPERCASE words are harder for them to read? Personally, I have absolutely no problem to read in any case and can read fluently upside down or even back to front reading from a mirror. Anyone have any ideas about this?
I speak only English, but I was taught to read phonetically, too; it works for a great many of the most commonly-used words in English. Exceptions (like the dreaded “gh” words) can be grouped and handled as special cases.
Most of the other non-phonetic words in English are actually foreign loan words; if you learn the rules of that language’s spelling (at least, the ones that pertain to those loan words) you can treat those phonetically, too (just with a different “set” of phonetics).
I read capitals phonetically, too…but I do find them harder to read because they’re TOO DARN LOUD. I mentally “hear” the all-caps as a kind of “shout.” Compare:
> No one will be admitted without a signed letter of permission.
> NO ONE WILL BE ADMITTED WITHOUT A SIGNED LETTER OF PERMISSION.
Doesn’t the all-caps version “sound” louder, even without an exclamation point? Reading a constant “scream” becomes exhausting very quickly.
Phew – I thought for a minute someone was going to start banging on about the added readability a serif font offers.
I also thought that someone would explore the differences of font applications and try holding this theorem up to blocks of text vs headline usages.
*HE WHISPERED*.
Because it slows people down, which forces them to read what is written, rather than skimming… it actually is so it is harder to read, sort of 🙂
Hard and fast rules about anything reveal a lack of awareness.
Genius, on the other hand, is evidenced by towering awareness.
Given that design brilliance is related most keenly to THE SITUATION AT HAND, I’ll take awareness and the talent to channel it into USEFUL THINGS over guidelines and rote knowledge any day.
Do we not have lead designers running circles around their less distinguished brethren by knowing exactly when and how to BREAK ALL THE RULES?
To the extent this article is an exploration of potential technique it is a success; to the extent it is a purveyor of dogma? USELESS…
I have worsening vision difficulties and it seems to me, CAPS are harder and slower to read. In addition, in emails, it does appear that someone is yelling. We have an on going trend in pharmacy practice to use Tall man letters to avoid drug errors when filling, translating or reading prescriptions. Although, this system is not practical everywhere, it has prevented many drug errors. So, my point is if one really wants to make themselves widely read and understood, one would want to write what’s easiest, less misinterpreted and fastest, despite of what one is “used to reading.”
Why is it OK to reduce contrast from maximum by coloring it gray instead of black, or sizing text to tiny and arbitrary px values instead of embracing every user’s preconfigured optimum text size (1rem)? Today’s web is full of the infuriating contrast reduction that makes two additional zoom levels beyond those required to bring text size up to the browser’s default and thus comfortable to read. By comparison, using all caps or not is a UX non-issue.