Can apostrophes swing both ways?

Adam Jagosz

I noticed that in some modern signage the apostrophe is given a somewhat atypical form.


At first, I found especially the first sample jarring. What's your opinion?
Fathoming the issue, I noticed how rare is font support for the reversed quote /uni201B. I thought it was a rather commonly implemented, though rarely seen in use, character (since it is placed beside the regular quotes in the Unicode plane). Is that because it is considered obsolete/deprecated/discouraged?

James Puckett

Most people don’t know the difference, so it doesn’t really matter.

Mark Simonson

In the first example, I think it was done for reasons of formal unity (echoing the curve of the S). I think it's actually okay because, even though it curves "the wrong way", the general orientation (NE/SW with the thicker part at the top) is correct.

In the second example, here again it seems to have been done for formal unity (notice the repetition of the same NW/SE angle throughout the design). This one is less successful and not very well drawn. The whole scheme here feels a bit naïve and the "backwards" apostrophe doesn't help. (See also Cutting Edge Logos.)

Adam Jagosz

Thanks for the link! I guess DeviantArt could be included in that blog, too.
I can see how the apostrophe in the Assasin's Creed logo fits neatly beside the S, but the character itself looks like an entirely new construct, not part of the original typeface (the concave top). And it's a tad too light, methinks.

Hrant Հրանդ Փափազեան Papazian

If you step back you might realize that the way type designers want the apostrophe (and the quotes) to curl is a bit OCD... Especially in logos there's so much room for fruitful play.

BTW does uni201B have a doubled version?

Michel Boyer

% grep 'QUOTATION MARK' NamesList.txt | grep HIGH
201B	SINGLE HIGH-REVERSED-9 QUOTATION MARK
201F	DOUBLE HIGH-REVERSED-9 QUOTATION MARK</code></pre><pre class="CodeBlock"><code><pre class="CodeBlock"><code>

Adam Jagosz

I recall seeing the reversed quotation marks somewhere, but sadly I cannot find any information about their use online. Wikipedia only shows early French usage of quotes that gave rise to guillemets:

I once thought that a reversed apostrophe would fit neatly after a /t, because in some fonts a regular one creates too much white space.

Hrant Հրանդ Փափազեան Papazian

Ah, those lovely marks I've come to call "quotemets"! They work marvelously – much better than our floating stuff: like guillemets they enjoy better spacing and avoid confusion with the apostrophe. See also: http://www.typophile.com/node/20061#comment-124405

But I had never heard that's how guillemets arose... And I don't think I believe it.

BTW, the flip side:

https://twitter.com/hhpapazian/status/760476106432450560

Nick Shinn

Flipped left quotes appeared in several pervasive American typefaces of the early 20th century.
I’ve always loved them in old movie title and cartoon lettering.
I made them the default in a couple of typefaces (language-tagged to English in Boxley).

Hrant Հրանդ Փափազեան Papazian

They also appear in the work of Carter and Licko. They make more sense. But instead flipping the right quotes makes even more sense.

John Hudson

Flipped left quotes appeared in several pervasive American typefaces of the early 20th century.

But cause problems in today's internationalised typographic world in which fonts need to support the conventions of multiple writing systems. Users in the European countries that use paired baseline (opening) and 'left' raised quotes (closing) really don't like this style.

Hrant Հրանդ Փափազեան Papazian

Language tag!

Nick Shinn

Right John, Verdana was corrected, IIRC.
That’s why I used a language tag.

**
Here is an interesting German writing style, from Drei Märchen by Wilhelm Busch, late 19th century.

Hrant Հրանդ Փափազեան Papazian

In my view, not "corrected". Like how Georgia's hybrid OS numerals were not "corrected" either.

John Hudson

Language tag!

It should be an option, but in practice it isn't a robust mechanism. One of the things that's lacking in OpenType Layout is an unambiguous implementation specification — heck, any implementation specification — that governs how characters with the Unicode script property 'common' get handled in terms of shaping engine paths, whether or not they are rolled into glyph runs with adjacent text, what happens with embedded languages/script and punctuation, etc.. Without such a specification, it's impossible to know exactly how to implement language system-specific variation for punctuation in a way that will produce reliable and consistent output.

Even non-language system-specific lookups affecting common punctuation characters can be problematic. For example, I found that in order to get a Stylistic Set variant form of the Indian danda and double danda punctuation marks to work in InDesign, I had to put the lookup under the DFLT script tag, even though the context and adjacent characters were Bengali.

Simon Cozens

An Australian advert designed by a French agency:

@hhpapazian Europcar Australia. pic.twitter.com/wWRdILrDz4

— Simon Cozens (@simoncozens) July 7, 2016

Kent Lew

There was a Cyrus Highsmith design that I worked on a few years back where he had drawn mirrored quotes. Matthew Carter relayed his cautionary tail against having that form as a default.

So, I attempted to work out a scheme to manage between mirrored and turned forms that would work appropriately in various language contexts. But because of the vast number of languages that use the different styles, it was impractical to try to address this through language tags and be confident of catching every possibility. As John says, in practice right now this still is not a completely robust mechanism.

In the end, I think we wound up taking a purely contextual approach. I’m pretty sure I chose to place it in {locl} — before any script/language declarations, so that it was included with all subsequent explicit declarations — rather than in {calt}, because {locl} is required and not turn-offable.

I can’t say this is the absolute best solution, but it seemed to work most reliably in our testing, and Cyrus was satisfied with this approach. We also included a Stylistic Set so that the turned quotes could be intentionally activated in all circumstances.



[Fun tip: In InDesign, if you have the “Use Typographer’s Quotes” preference turned on and you set the language attribute on a selection of text, it will apply what it thinks is the preferred quotation style for that language when you type “dumb” typewriter quotes.]

John Hudson

Kent, what was the nature of the contextual substitutions? I can't imagine how this would work with any reliability beyond a single word length (and with some layout engines wouldn't even work for a single word, because quote marks are not reliably rolled into glyph runs with adjacent letters).

Michel Boyer

There is a problem with French, Russian and Ukrainian: there should be no single quote, just the chevrons. Probably with Slovenian too.

Hrant Հրանդ Փափազեան Papazian

People do all kinds of things with quote marks. Even using guillemets in English. Thankfully.

Christian Thalmann

Michel Boyer said:

There is a problem with French, Russian and Ukrainian: there should be no single quote, just the chevrons. Probably with Slovenian too.

That may be to indicate that single quotes are used as the second level of quotation? Dunno.

Michel Boyer

That may be to indicate that single quotes are used as the second level of quotation? Dunno.

In Russian, according to https://ru.wikipedia.org/wiki/Кавычки, the second level is given by „…“. Same thing for Ukrainian (according to that Russian wiki). In French, according the French wiki, there is a number of choices but none with single quotes.

Kent Lew

That may be to indicate that single quotes are used as the second level of quotation? Dunno.

Sort of. Those single quote marks are indications of what InDesign thinks it should do with a single typewriter quote to become a “typographic quote” in those languages. I’m not vouching for it.

Of course, that isn’t the same as saying that a single typewriter quote should be used for prompting a secondary level quotation. Sorry if that was misleading (it was an old graphic I dug up).

John — I’d have to go back and look up exactly what I did, which I don’t have time for right now. But, as I recall, the basic logic was that if a quoteleft/quotedblleft is preceded by anything except some manner of space, then it can be assumed to be a closing mark, at which point the turned form is substituted. Maybe an oversimplification, but seemed satisfactory for a display face like that.