Matthew Peterson | Essays

The Cuckoo Bird and the Keyboard

Easily the most maligned key on your computer’s keyboard lies just to the left of “return” and represents what appear to be single and double quotation marks. It is a cuckoo’s egg in the designer’s nest. It doesn’t belong. Unless we put a finger on the key’s history and purpose, we risk being cuckolded, in the truest sense of the word.

Graphic design, as a trade, is remarkable for the scope of its jurisdiction, ranging from the first-read of a designed object down to structures of meaning and fine editorial details. But we neglect the latter when favoring the former in our discourse. An unbalanced infatuation with posters and book covers delimits our collective perception. This wide-angle view precludes any awareness of the influence exerted by a small character displayed on our keyboards.

Most of us would agree with this statement: The quote-like characters on the keyboard are not actually English quotes; they are primes, and serve as foot-inch indicators. Trained designers “know” this, but it is only half true (and half false, and in that order).

The True Half: “The quote-like characters on the keyboard are not actually English quotes...”

The characters displayed on the key in question [ " ] don't have a singular name — there is no consensus. They’re alternately referred to as typewriter quotes, dumb quotes, ASCII quotes and neutral quotes. While typewriter quotes, as a term, is quite revelatory, neutral quotes (or simply neutrals) is favorably concise.

Designers are famously nauseated by novices’ use of neutral quotes in place of true quotes. They are poor surrogates. To the trained eye, they look remarkably out of place. There are two reasons for this. The one is that our eyes have been trained: the point is made repeatedly in a designer’s education by sensitized instructors whose concern is correctness. The other is simply that, inherent in the actual typeface designs, neutrals often are out of place.

As a striking example, the neutrals in the digital font Franklin Gothic ITC BQ have formal characteristics found nowhere else in the type design. They consist of a rounded top, a straight taper and an angular bottom. Roundness within such a small form is completely unprecedented when considering other characters: the period, comma, quotes and accents are all completely angular. These neutral quotes engender a conflicted aesthetic.

The reason for this seems obvious when considering other digital versions of ITC Franklin Gothic. The neutral from a Microsoft version supplied with PCs has a full taper with no rounded top; Adobe’s has a tapered bottom below a heavy rectangle, all angular; and Bitstream’s is a plain, thin rectangle. Aside from this one odd character, these typefaces show only the subtlest of differences. All of these Franklin Gothics are based on the same ITC source design, which predated the computer. The differences are easily explained if the neutral quote wasn’t in the originals. If added by another designer in another era (a different designer for each digitized version), the incongruities would make perfect sense.

Reproductions of phototypesetting films in Jim Craig’s Phototypesetting support this hypothesis. These films predate desktop publishing (thus digitization of existing typefaces) and lack neutral quotes altogether.

So where were the neutral quotes before the computer if not in professional printing? I believe the answer is in our discarded but descriptive term typewriter quotes, for they first appeared in the 19th century on typewriters. The reason for their invention was purely economical: space was at a premium, and separate keys for opening and closing quotation marks [ “ ” ] seemed wasteful. The economic imperative was strong; for example, some typewriters omitted the numeral “1,” surely to be replaced in practice with the lowercase “l” or capital “I.”

For a singular mark to supplant both opening and closing quotes, a new form was necessary, one that favored no direction and appeared reasonably familiar.

So neutral quotes were born unto the typewriter. Since typewriters were a direct writing method and couldn't mass-produce, the neutral quote was confined there. As only proxy for true quotes, neutrals weren’t transportable to professional printing.

This changed with the personal computer and desktop publishing. Previously, input with a keyboard was performed live onto paper by a typist. With desktop publishing, keystrokes on keyboards — now wired to computers — were instead saved and directly output to professional printing. This is unfortunate, as the neutral quote — that wholly degenerate form — has since been utilized so relentlessly.

The neutral quote is nothing. Considering its purely derivative nature, it should serve no function whatsoever in proper typesetting. Perhaps we should know this, but we don’t. It so happens that there is a use to which we commonly ascribe the neutral quote.

The Half-Truth: “...they are primes, and serve as foot-inch indicators.”

The second half of our statement has some truth to it: prime symbols [ ″ ] are proper abbreviations of feet (single) and inches (double). They can also signify minutes and seconds of time or arc, and have further uses in mathematics.

Where the statement goes wrong is in misidentifying neutrals as primes. We’ve learned that neutrals are merely degenerate quotes, only recently liberated from the typewriter. But primes were serving their numerous functions long before neutrals emerged.

Surprisingly, typographic manuals — authoritative sourcebooks for designers — frequently confound primes with neutrals. I’ve found only one of the more celebrated manuals that correctly identifies primes: James Felici’s Complete Manual of Typography (which also details the true pedigree of neutral quotes). But why are many popular textbooks erroneous and only one accurate? Perhaps it’s too easily assumed that a prominent character on the keyboard must be something, just as an egg must belong in its nest. The prime — more than any other character — seems a good candidate for misattribution because it’s otherwise hard to find and does bear a superficial similarity to the neutral quote.

But similar is not selfsame. Programmers know this because in functioning code, a prime and a neutral quote are precisely as discordant as any two unrelated characters. As such, it isn’t surprising that the source of my own neutral-ain’t-prime realization came from a programmer's HTML tutorial, and not the one correct type manual.

No, it’s still surprising. This issue is undeniably under the typographer’s purview. I haven’t seen a single correct use of primes in any printed ephemera in the last few years (including my own recent work, I’ll admit). But art publishers — a specialized industry listing canvas dimensions well before the neutral even existed for printing — quite often get it right.

Canvas dimensions often include the multiplication symbol [ �— ] in addition to primes—or they should. But typographers often use the letter “x” instead. Like primes, the multiplication symbol is not represented on the keyboard. Like neutrals, the letter “x” does appear on the keyboard. But there is a noteworthy difference between these two right-wrong pairs: the letter “x” was never exclusive to the typewriter. It has always been available in professional printing. The “x” error occurs as often as it’s avoided, regardless of publication date. This suggests something about earlier (and correct) use of primes. It seems to be less about some lack of competency in the present, and more about a lack of opportunity to err in the past.

Though an obsessive attention to detail might eradicate the “x” error from practice, that’s not enough for the neutral error. It will persist so long as it’s supported in typographic manuals. When experts’ source material contains mistakes, they tend to be amplified throughout the practice.

A fresh sensitivity to primes is further hindered by their chronic absence in most digital fonts. The prime appears almost exclusively in a few core system fonts for Macs and PCs — supplied to meet the more exacting needs of the programmer (however counter-intuitive that seems to a typographer). They are exceedingly rare in fonts marketed to professional designers, which compounds the problem: typographers aren't about to employ characters missing from their typefaces.

The design community has shown concern over the increasing availability of design-based tools, worrying that the novice who makes design is a threat to the profession. They’re not. Indeed, they’re part of the profession — by definition — when designing, are they not? But however wide your pronoun (we) casts its net, the truth remains: Proprietary knowledge will always separate the expert from the novice, so long as it’s retained.

This essay is not so much about the cuckoo bird or its egg. It’s about the caretaker: a reed warbler, let's say. Will the reed warbler return to its nest and notice something hidden in plain sight, and suspect that it's out of place, and question its presence, and care, and act accordingly?

The cuckoo bird and the keyboard, in their little, discrete way, demonstrate that our tools not only affect how we work, but what we know.

They also suggest that punctuation marks are the stuff of perfectly valid concern. To establish expertise, we must be attentive to — and concerned with — such minor details as these. To maintain expertise, we must identify fallacies in our practice. And then, with pleasure or dispassionate handicraft, reject them from our nest.

Posted in: Business, Technology, Typography

Comments [36]

|| The prime appears almost exclusively in a few core system fonts for Macs and PCs — supplied to meet the more exacting needs of the programmer (however counter-intuitive that seems to a typographer).

I don't know; I've never, ever needed a "prime" character as a programmer. Programming langauges (these days, at least) tend not to require characters that aren't directly available on (en-US) keyboards.
Adam V.

Not all programmers, of course, but surely many in certain areas. On the Mac, we have Lucida Grande, which the system uses and which contains a mindblowing amount of glyphs, in order to accommodate all possible needs. (The PC equivalent might be Arial Unicode, but I can't say.) What I'm considering as "more exacting" is that from a programming perspective a character looking kinda like another is meaningless... If it's not available when it needs to be, that's a problem. Some programmers had to ensure there were as few problems as possible in the system's operation. (But for the record, I don't know programming!)

Now if we can just teach all the major U.S. news networks and local TV news stations to simply use real quotes on screen… (Who does that work? Interns from journalism schools? Is there an association of TV news graphics people?)
Maybe AIGA, SOTA, ATYPI, Adobe, et al could contact all the manufacturers of on screen TV graphics generators and ask them to make it automatic - like it is in most word processing and page layout software…

Look at the Typographer's Quotes preference option in Adobe InDesign C3.
Carl W. Smith

It's interesting how you go to the trouble of getting the quote marks right in your article, whereas on the rest of this site neutral quotes are used, both in entries and in the observed feed, including the one that points to Apostrophe Atrophy.

Does that say something about how internet browsers should interpret quote marks (auto smart quote conversion?) or does it say something about how neutral quotes have managed to change the orthodox view of how quotes are used via the computer?
Bronson Fung

"...our discarded but descriptive term typewriter quotes, for they first appeared in the 19th century on typewriters. The reason for their invention was purely economical..."

Yes, and/but there's the interesting channel through telegraphic alphabets: International Morse had one symbol for "inverted commas (quotation marks)", but also had a symbol for a true multiplication sign.

The five-unit telegraphic alphabet No. 2, based on Murray (and still earlier Baudot) code, had no double quotation mark whatever, nor multiplication symbol. Keep in mind that there were 2x2x2x2x2 equals 32 combinations of which two were allocated to letter and figure shifts. The spaces available in figure shift mode, after 0-9, were given over to various other symbols — ?:%@., etc — and to a couple of "stunt" signs, e.g., "who are you", "bell".

ASCII (alphabet No. 5 for data communication and telegraphy) allocates a position for double quotation marks, and other symbols like "greater than" and "less than", but no multiplication marks. ASCII's creation in the 1950s involved battles between the communications (telegraph) and data (computer) communities, and would eventually give way to Unicode.

The point of this footnote is to bring the telegraphic alphabets (which were present on teletypwriter and teletype keyboards) into the picture. I imagine (but don't know the degree to which) the Linotype perforated tape code related to any of these, but suppose there was some connection.

Even today, I use tend to use dumb quotes for webpages, because I've no confidence that people's browsers are set to show & l d q u o ; or its terminal twin.
John McVey

...even our favorite brands occasionally get it wrong as well.
Doug Bartow

re: favorite brands,

No wonder the NYT missed that typo — albeit the advertiser's — so preoccupied are they with improvements to their opening pages.
John McVey

I found this blurb on the micorsoft typography developer page.

Cute childish names like 'smart quotes', 'fancy quotes', 'curly quotes', '66 and 99' or 'duck's feet' are some of the terms used today to refer to these typographically specific characters. Fonts in early letterpress typography had no quotation marks. A typographer would use a comma, shifted or rotated as a quotation mark. It could be said these marks are traditionally based on the comma design. These methods were inadequate and design specific quotation marks were developed.

Anyway, I just find it more interesting that the French, Germans, Greeks have similar dealings with guillamets, angle brackets, and the mathematical operation signs as well. NOW: <> Could be this stuff all originated because when new technology comes around it is rustic in the beginning. Someone takes something available or existing and just turned it upside down or around to form another. Of course, how calligraphy and books were written and stone were cut...

Well, there must have always been the divide between the mathematically minded and the literature minded, and funny how some can serve both or just crossdress when they need each other. Nordics meeting the greek or latin. Or some other more ancient drang or race to the rites of being written right.

Thank you for this article. I teach this to my typography students as well, but I was beginning to think I had made it all up since I see "typewriter quotes=primes" in type book after type book.

Why are primes not included in more typefaces? It seems like such a small added step to take when compiling all the pieces of a typeface into a font.
Troy Johnson

I don't particularly like having angled quotation marks or apostrophes, nor do I attribute an 'orthodox' rightness or wrongness to them, but I find this is a fascinating post nonetheless, as I enjoy these precise expansions on a deceptively simple theme.
M. Nestor

I enjoyed Bronson Fung’s comments. Maybe it’s just my browser, but it seems there are inconsistent apostrophes even in this article — see the second sentence in the third paragraph from the bottom: ‘It’s about the caretaker: a reed warbler, let's say.’ If you have to nitpick, I would expect Design Observer to practice what they preach. BUT, if it is indeed a browser issue, please disregard my petty remarks.
p. byer

"Even today, I use tend to use dumb quotes for webpages, because I've no confidence that people's browsers are set to show & l d q u o ; or its terminal twin."

I keep hearing this crap about how it's a browser problem, people's browsers need to be set up to support correct marks, etc.

There is no browser in existence that can't handle Unicode at this point. I'm serious. It's 2008. I don't know where that excuse is coming from. Internet Explorer 5 could handle smart quotes just fine.

sigh. thanks Werson. will continue to use dumb quotes, then, just because I'm dumb.
John McVey

After reading that the differences between neutrals/primes and standard quotation marks are not outlined in most type manuals, I immediately reached for my well-worn edition of Bringhurst's "Elelments of Typographic Style" that sits about three inches to the left of my monitor.

I was suprised to discover that, indeed, 'ole Bobby barely scratches the surface of this particularly fascinating debate, though he definitely takes a very clear stance of the use of × versus x and other common typoraphic quandaries. It got me thinking, if I didn't learn to be ferociously anal about quotes versus primes from the Bringhurst, then where did I pick it up from?

It turns out, both the Associated Press style guide and the Chicago Manual of Style contain proper usage guidelines for these pieces of punctuation, having pretty much memorized the AP guide during a brief stint as a journalism major back in college. Really fascinating, as I consider this more of a design-related specification than a grammatical one.

Even more interesting that the mistake apprears so fequently on broadcast news programs, which I too have noticed, being that you would imagine a bunch of journalists and copy editors would get that one right.
Darrin Crescenzi

"sigh. thanks Werson. will continue to use dumb quotes, then, just because I'm dumb."

do it because you're lazy like the rest of us =D

Werson and McVey should both realize that you don’t need to use character entities for proper quotation marks in HTML; if you declare Unicode or a similar compatible encoding, you can and should just type the characters. They’re just quotation marks, for G-d’s sake. There’s nothing that special about them.

Now, if you’re running Windows, there is almost no chance you will know how to type them. This would be your problem.
Joe Clark

Does it really matter?

Yes it matters, douche.
Kula bácsi

|| The prime appears almost exclusively in a few core system fonts for Macs and PCs — supplied to meet the more exacting needs of the programmer (however counter-intuitive that seems to a typographer).

I don't know; I've never, ever needed a "prime" character as a programmer. Programming langauges (these days, at least) tend not to require characters that aren't directly available on (en-US) keyboards.


open(FH, "somefile.txt") || die "Cannot open file: $!";

A fun way to inform the masses: Link

We Web designers have been trained to manually replace all ampersands to create valid XHTML. Bullets are manually converted into unordered lists. The markup is structured cleanly and semantically. And just for visual formatting, we manually replace all em-dashes and en-dashes with the correct XHTML entity. But quotation marks get lost in the mix. Why? Because it takes way too long to fix. We are a lazy bunch and understand that if it's published through HTML, the rules of style have relaxed somewhat, in keeping with the technicalities in how online content is published. To keep pace with a glut of content, to a certain extent, you have to just let it flow. And if you disagree with this, you're being old school and you don't fully understand the problem.

True, if the source document contains the correct quotation marks, then we, as good Web designers, should insert the corresponding entity (& l d q u o ; and & r d q u o ; [without the spaces]) into the XHTML. But, as many of you know, that is actually a big pain; there's never enough time. There are different kinds of quotes in an HTML document and once you've pasted the article into your text editor, you can't globally search/replace every neutral quote with the correct one because it could possibly screw up the entire document's formatting, and so it would have to be a manual process. This has been happening for years. When you're marking up pages upon pages of content on a deadline, the correct quotation mark invariably gets lost – because we realize culture is changing to accommodate the rapid pace of online publishing and our Web design laziness.

By the way, if anyone believes browsers should interpret the neutral quote mark as anything other than ", then enjoy your WYSIWIG editor. I, for one, feel annoyed when browsers render my markup in a way other than how I explicitly mark it up. But I think I could get used to some kind of intelligent quote replacement.

A quick look at the windows character map would tell you that a key stroke of Alt+0147 would produce the opening quoatation mark (“) and a key stroke of Alt+0148 would produce the closing quotation mark (”); furthermore, these characters display correctly in browsers, thus, rendering the use of & l d q u o ; etc absolutley pointless.

There is no reason to be lazy as a web developer, it’s the small things, like the correct use of characters, that makes a design stand out.

Surely any designer, website or otherwise, worth their salt knows this?



Carl, I completely agree with you that it makes for a better reader experience to display the correct quotes. However, in regards to the statement "these characters display correctly in browsers"...this is actually wrong. If you're writing HTML, you cannot and should not use those keystroke characters to generate the quotes. It is NEVER valid HTML to insert quotes that way. You even end up with funky characters under certain DTDs (note: ISO-8859-1 renders them properly but it still doesn't validate). This is a much worse way to approach the problem than just settling for the neutral quotation mark. Conscientious designers know this, and that along with including details such as the correct right and left side quotations, another mark of a great design is one that was also rendered from valid HTML.

Another point is, from a purely practical standpoint, if it doesn't bother you to write invalid markup [ugh] and you are designing and developing a website with a lot of content, you definitely aren't going to enjoy doing all those keystrokes by hand. You also won't enjoy inserting the correct XHTML entities & l d q u o ; and & r d q u o ; to make it valid either. Because after you paste in content, most text editors replace the invalid left and right quotes with their always-valid neutral quote counterpart. And since there are different kinds and uses for quotes in an HTML document, globally searching/replacing every neutral quote with the correct one could screw up formatting for the entire page. Therefore, it has to be a manual process...and often, tragically, it just doesn't get done. Think about a site with hundred of press releases getting transferred to a new template. It actually isn't realistic to manually correct all of those quotes. Furthermore, I think culture has changed - many readers tend to gloss it over and accept it, because it's been incorrect for so long. Not that that's a good thing.

I could get used to some kind of intelligent quote replacement if it were implemented well in Web browsers. Like how we don't have to use two spaces after a period anymore.

Aquarium fish freshwater tropical
Adding water to your fish aquarium

Try running the code below through the validator over at w3c, it validates. Tested this in IE7, Firefox, Opera etc and it works fine.

In regards to “You even end up with funky characters under certain DTDs”, if you care about valid markup (as you rightly do) then there is only one DTD you should be using and that is the one used below.

I have just pasted quotes into three different text editors and the quotation marks stayed intact; perhaps you have just been unlucky.

<?xml version="1.0" encoding="ISO-8859-1"?>
<!DOCTYPE html
PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
<title>Virtual Library</title>

<style type="text/css">





<p>“Valid xhtml with quotes”</p>


Carl, I stand corrected.

This is very interesting. I didn't know the difference between primes and neutrals myself. In fact, I still don't. It's an excellent article, but it doesn't explain how the primes and neutrals differ and how you would tell them apart. Would someone enlighten me?

I found this article to be pretty interesting. As a programming student, I have never needed to use a "prime" character, or any character for that matter that isn't offered on the US-EN keyboard. However, in my work in web-designing, I have found that copying and pasting a needed prime into my code. I have never needed to run a code that will allow me to use the primes needed, however I do know from experience that the above code is valid, and has benefited when using primes in some cases- I have had some compatability issues with Opera explorer before.

Forgive me if someone already posted this, but I found this article too, on the same subject...

Joy Gallucci

take prefer to commiserate with with BBcode :( for me does not revolt out to spawn references to the gripping pictures and video on the topic. Or regurgitate down fasten together on this topic

Kula bácsi, great insight. At least I know, that in a world going to sh*t, you will be there triumphantly using proper quotations and hopefully taking the lives of those foolish enough to use primes.~


Nonagon is quite incorrect and is confusing data and markup. For the umpteenth time: Declare a Unicode character encoding for your file and just type the characters. You don’t need entities and quotation marks are not special in any way. If you receive a file that contains correct quotation marks, dashes, and the like, you mark up that file; you don’t convert its characters to entities.

Again, people: No matter how many times you state that real quotation marks absolutely have to be converted to entities and it’s just too much trouble sometimes, you won’t make either of those facts true.
Joe Clark

@ dotdotdot

Kula bácsi, great insight. At least I know, that in a world going to sh*t, you will be there triumphantly using proper quotations and hopefully taking the lives of those foolish enough to use primes.~

So, “in a world going to sh*t” (sic), you:

a) took the time to read this article
b) took the time to comment it

but anyone caring about “proper quotations” (sic) is to be damned.

Well, when you asked if any of this matters, you never added “compared to helping save the tropical rainforest” or “compared to fighting world poverty” (as —I am sure— you do).

People smarter than me have argued that the world going to rot has something to do with people not caring passionately for each and every detail. Something like cultural entropy.

And, yes, typography is a large part of our cultural heritage. Convenience only buys you a culture of McDonald’s.

(this comment of mine is in no way affiliated to the previous commenter Kula bácsi)

One afternoon, I was in the backyard hanging the laundry when an old, tired-looking dog wandered into the yard. I could tell from his collar and well-fed belly that he had a home. But when I walked into the house, he followed me, sauntered down the hall and fell asleep in a corner. An hour later, he went to the door, and I let him out. The next day he was back. He resumed his position in the hallway and slept for an hour.
This continued for several weeks. Curious, I pinned a note to his collar: "Every afternoon your dog comes to my house for a nap. "
The next day he arrived with a different note pinned to his collar: "He lives in a home with ten children - he's trying to catch up on his sleep."

I cried from laughter
Sorry, if not left a message on Rules.

A few comments:

1. Left quote marks were originally set in metal foundry type by rotating the comma through 180°, which is why they rode a bit low in some faces.

2. A character which changes its shape depending on contextual position? That used to be the case with "s" up until about 1800, with short s at the end of words and long s elsewhere.

3. More egregious than a neutral character representing both left and right quote marks is the widespread use of single left quote marks as apostrophes in abbreviations such as years, e.g. ‘08, caused by "smart" layout applications. This incorrect usage now appears to be the norm.

4. Why aren't neutral quote marks simply angled lines? Wouldn't that make sense in sans serif faces? That way, they'd look much more like proper quote marks, and also primes.
nick shinn

I'm not sure if someone has posted this or not, but this web page shares some interesting technical information on these glyphs. :)

Louisa Nicholson

Jobs | July 18