Proper placement of exclamdown and questiondown

speter's picture

A question to our Spanish language typographers: where vertically should exclamdown and questiondown (i.e., ¡¿) be positioned? I've seen examples with the characters essentially on the baseline, and I've seen examples with the characters shifted down, though never as far as the level of decenders.

Ricardo Cordoba's picture

Hi there, Steve!

They should be placed shifted down, but not as far as the baseline. Some (most?) typefaces have them aligned at the ascender by default, which is also wrong. I can't post images just now, but I will try to do that tomorrow, if that helps and if you can wait that long.

speter's picture

¡Hola, Ricardo!

Yes, images would be great! (No rush. I'm working on a typeface and wondered what to do when I was setting some sample text. Too far up, and they looked odd, and too far down, they looked odd in a different way.)

Ricardo Cordoba's picture

Here's an online example... You'll notice that the opening exclamdown is lined up at the x-height.

speter's picture

Thanks, Paul! Very interesting...

hrant's picture

Two less-obvious things to consider here:
1) The inverted symbols are usually followed by a cap.
2) Mind the confusion with the "i"!

hhp

paul d hunt's picture

i think most current typefaces use the x-height alignment as a default and have another case sensitive version for all-caps settings which sits on the baseline.

k.l.'s picture

R.C. -- Some (most?) typefaces have them aligned at the ascender by default, which is also wrong.

H.P. -- 1) The inverted symbols are usually followed by a cap.

Hello Mr Cordoba, how wrong is this? I made the same observation as Mr Papazian, and for this reason -- deliberately -- chose the position between baseline and H-height. Another issue is that an x-height-aligned questiondown may require quite a few kerning pairs.
But seeing '¡Hola, Ricardo!' above (exclamdown aligned with uppercase), I could as well argue that descending exclamdown and questiondown are easier to recognize. Hm.

ebensorkin's picture

Like all things in type the details of the face ( style, weight, serifs etc.), and your intent will come into the equations as well. The shape of the Question mark & Exclam are important too.

Not that it is important, I tend to like the feeling of x-height ish alignment better. But I don't read it very much so my opinion is only worth just so much. Some ideas about why I feel that ways include:

- It might be that lowering the mark before a capital let's me see the the next letter faster/easier. eg - the mark gets in the way of my reading less. (If true this would be a good reason or an endorsement )

- It might be that a lowered x-height ish alignment has a better 'cue value'. Or put another way it may be that I can recognise the ¡ or ¿ faster when they take that alignment. (If true this would be a good reason or an endorsement )

- It feels more even more 'different' than an H height alignment & I like the exotic quality ( not an endoresement or a negative )

None of this is more important than using your own Eyes. Find what works best for the face & the reader.

Ricardo Cordoba's picture

Eben, I was just saying the way it is. This is what I was taught whenever I had Argentine schooling, and what I see in most of the books I have in Spanish, if they are set in Spanish-speaking countries.

The inverted symbols are usually followed by a cap.

Well, if they are at the beginning of a sentence, yes, they would be followed by a cap. But a question or exclamation can begin mid-sentence, thus:

Hrant, ¿qué opinas de esta fuente?

Hace calor, ¡quiero un helado!

These upside down symbols act as visual cues, to let a reader know when to change intonation, much as the accents (or tildes, as they are known in Español), let you know where the stress falls in a word -- but that is a whole 'nother subject. :-)

Ricardo Cordoba's picture

Hello Mr Cordoba, how wrong is this?

Well, I just read the Microsoft guidelines that Paul posted, and I didn't know about the display setting... I don't have too many examples of that with me, but I did find one, and sure enough, the upside down exclamation point is aligned with the rest of the caps... I'll try to scan this and some text face examples tomorrow, and post them here.

Another issue is that an x-height-aligned questiondown may require quite a few kerning pairs.

Hey, that's our job as type designers anyway. Most punctuation needs to be kerned. Quotation marks can be annoying, but you don't go around changing their vertical position to make them easier to kern... right? ;-)

Ricardo Cordoba's picture

Here's another online example. Malena, a typeface designed by Felix Lentino, a typography prof at the University of Buenos Aires.

k.l.'s picture

Thank you!

R.C. -- Hey, that’s our job as type designers anyway. Most punctuation needs to be kerned.

Definitely. Yet one cannot kern everything, I mean, even in high-quality typefaces I find combinations which slipped attention but may occur in this or that language. So if adjusting shapes helps avoid necessity of kerning in the first place, great.

E.S. -- None of this is more important than using your own Eyes. Find what works best for the face & the reader.

The latter in particular, since I do not read Spanish! It would be interesting to hear many opinions of Spanish speaking collegues.
Just like in previous discussions (would Americans accept guillemets? what about spaced punctuation marks?), I am interested in finding out whether something is de facto convention, or readers do not really care but typographers are taught or teach so, or is propagated by specific typographic 'schools', or is the opinion of individuals.

speter's picture

It would be interesting to hear many opinions of Spanish speaking collegues.

Karsten, that's exactly what I want. I've been looking at as many Spanish-language sources as I can for actual use, but the reasoned opinions of a few real Spanish-language typographer would be very valuable.

would Americans accept guillemets? what about spaced punctuation marks?

Yes to the first, no to the second (meiner Meinung nach).

Ricardo Cordoba's picture

Definitely. Yet one cannot kern everything, I mean, even in high-quality typefaces I find combinations which slipped attention but may occur in this or that language.

Karsten, I agree. And one cannot anticipate every situation, especially in languages one does not know! (If I remember correctly, The Elements of Typographic Style is a good resource, as it gives examples of frequently occurring letters or combinations for several languages.) And then once the typeface is out there, different typesetters will have their own kerning preferences... (And I just saw your e-mail, which I'll reply to!)

I am interested in finding out whether something is de facto convention, or readers do not really care but typographers are taught or teach so, or is propagated by specific typographic ‘schools’, or is the opinion of individuals.

All good questions, sir!

...but the reasoned opinions of a few real Spanish-language typographer would be very valuable.

Paging Ale Paul... Eduardo Manso... Miguel Hernández... etc.

:-)

Ricardo Cordoba's picture

OK, Steve, here are some scanned examples from Spanish-speaking countries. I'll start with text faces and then move on to display faces...

These first two images are from the novel Puras mentiras, by Juan Forn, an Argentine writer, and published by Alfaguara in 2001 -- presumably this book was typeset in Argentina (sorry, the colophon is rather skimpy on details). Here the exclamdown and questiondown line up at the x-height.

Ricardo Cordoba's picture

These next two examples are from the book Del Diseño, by Yves Zimmermann, a graphic designer from Barcelona. This volume was published by Editorial Gustavo Gili, of Spain, in 1998. (Again, there is no colophon explaining who did the typesetting or which faces were used -- usually, books published in Spanish-speaking countries simply have a note at the end stating the printer's name, how many copies were printed, and the date. Sometimes the designer of the cover will be mentioned on the copyrights page.)

Anyway, these examples are interesting because one uses small caps, but the questiondown still lines up with the small caps rather than the initial cap; the second example shows the questiondown following a colon -- and still lining up with the x-height.

Ricardo Cordoba's picture

Oops, it's rather late -- I'll post the display face examples tomorrow.

speter's picture

Very nice examples. I've been looking at all the Spanish-language editions on newseum every day*, and I think there's a definite preference for x-height alignment. Nearly all the uses I've seen there (admitedly very unscientific methodology) have caps-height alignment only in ads (which seems to me to connote shouting).

(Del Diseño looks like it could have used better typography: there's no fi-ligature in manifiesta, and those small caps look faked!)

*Why this question has to obsess me when I am in Moscow, rather than when I'm back in New Jersey is beyond me...

ebensorkin's picture

I think the design of an individual typeface cannot be completely hostage to convention / tradition any more than it can afford to ignore them. Sometimes it's a great idea to break a rule. Not often. The only way to know is to look; and then weigh the consequences for yourself.

So it could be that those with the H height are being somehow culturally insensitive or something but - I can't quite sign up for that because I am not utterly sold on the idea that there is a 'convention' yet. Partly I get this from looking around the font universe.

Looking at News Gothic the Bitstream version uses the H height. The Linotype NewsGothic uses the x. These seem to be true house-wide ( mostly ). Freight uses the H. Bello, & Feijoa the x. Tiptoe & Omnes the H. Amalia, & Fedra the x. Parry & Sansa Slab the H. Candy Script ( and all of Ale Paul's others) use the x more or less. Eduardo's fonts ( emtype ) all seem to use the x.

It also seems to me that the heavier the face is, and the more geared to display work the face is, the more an H height alignment seems reasonable. The spirit of the type might suggest what is right too. A slab Egyptian or a modernist face 's 'raison d'être' might be most compatible with an H height aligned '¿'.

Still, Ricardo, thanks for posting the link to Malena. And your example is exactly the kind of thing that looks pleasant/natural to my foreign eyes; naive as they are.

I also agree that hearing from further from Ricardo, Ale Paul, Eduardo Manso, & Miguel Hernández would be far more useful than the meagre theory & observations I have offered.

Still, since that's all I have I will go on a little further.

It seems to me that the design of the ¡ ¿ ? & ! probably aught to be done differently if you intend a heavy Spanish/Catalan use. In addition to certain alignments being easier to kern for; or looking better; ( or insert your concern here )- certain forms probably work better when they are spun around than others. Maybe softer hooks? Does anybody create a different shape for their '¿' vs. '?' ? It seems like that might be a very good idea as well. I should check to see if that is already being done actually... Maybe these are silly ideas. What do you all think? Any examples?

kentlew's picture

Both of Ricardo's examples are set in Adobe Garamond, so they don't provide much variety of evidence. I would guess that, in these cases, the position of the questiondown is more a matter of the typeface/font design, than any typesetting convention. Presumably, however, Adobe and Slimbach did some surveying and gave some consideration to the standards that they've adopted for their Originals series. And, presumably, the designers of these books gave some consideration to what the questiondown looked like before settling on Adobe Garamond. (Or not.)

What these examples do show is that the x-aligned questiondown does not violate any general cultural acceptance.

The best perspective, however, will still be provided by a native Spanish-speaking typographer or type designer who is steeped in his/her own tradition.

I would hypothesize that historically (meaning handset metal) the questiondown was originally just an inverted question mark, and so it's position relative to the baseline would depend upon the proportions of the font at hand. (Similarly, the earliest opening quote was an inverted comma, and so the alignment of opening quotation marks sometimes varied from that of closing quotes/apostrophes.)

Relatively longer descenders (like in a Garamond) would yield a top alignment closer to the x-height. Relatively shorter descenders would yield a questiondown that aligns higher.

-- K.

ebensorkin's picture

Nice points about proportion Ken!

Ricardo Cordoba's picture

Eben and Kent, thanks so much for your thoughtful and articulate observations!

Eben said... I am not utterly sold on the idea that there is a ‘convention’ yet...

Kent said... What these examples do show is that the x-aligned questiondown does not violate any general cultural acceptance.

I need to expand a little on my earlier comments... I had some schooling in Argentina and some in the States... It is in Argentine grade schools that I was taught to vertically dip the upside-down exclamation point and question mark in my handwriting... After a while it becomes second nature and you don't even think about it. Of course handwriting is not the same as typography, but they are closely related, no? :-)

Both of Ricardo’s examples are set in Adobe Garamond, so they don’t provide much variety of evidence.

Thanks for catching that, Kent! As tired as I was two nights ago, picking out books, and as tired as I was last night, scanning and posting examples, I can't believe I didn't even notice that either of them were set in Garamond to begin with!

Now, I don't have that many books in Spanish with me right now; I left a lot of them in Buenos Aires when I moved to New York. But even if I had them all here, I suspect that a lot of the fiction titles would also be set in Garamond... While I was working in Buenos Aires during the 90s, there was a pretty big economic recession in progress, and I know that most book publishers could not afford to spend a lot of money on cover design and illustration... I doubt that they would have been able to invest in many typefaces, either... (I could be very wrong about this, though. It would be good to hear from some other Argentine designers, to see what they say.)

Kent also said... I would guess that, in these cases, the position of the questiondown is more a matter of the typeface/font design, than any typesetting convention... and ...presumably, the designers of these books gave some consideration to what the questiondown looked like before settling on Adobe Garamond. (Or not.)

Very nicely put -- maybe they did, maybe they didn't.

I'll be posting some display face examples next, and they may just add to the confusion!

Ricardo Cordoba's picture

Here are a couple of scans from the book ¡Yo no fui!, by Argentine cartoonist Quino (first published by Ediciones de la Flor in 1994; layout and cover by Pablo Barragán).

In the first image, from the cover, the exclam-down lines up with the caps... is it the typeface or the typesetter?

And the second image is from one of Quino's cartoons... I decided to post it in light of some of the topics coming up in this thread. I know that Quino studied architecture at university, and that he is oh-so-meticulous in his drawings... If you look at the bottom lefthand drawing, the two exclamatory sentences in a row provide another good reason to dip the exclam-down. Also notice how Quino makes his exclamation points slightly diagonal, to either give emphasis or to keep the exclam-down from looking like a lowercase i...

ebensorkin's picture

I really enjoyed looking at these. If nobody shows up I may send the URL around to see if we can get additional participation. Maybe Carl Crossgrove can comment about whether there is an in house standard at Mono/Lino as well. :-)

Ricardo Cordoba's picture

Here are three examples from a children's magazine, Billiken -- I briefly (very briefly!) worked there as a layout artist in 1992, and this was before they switched over to Macintosh production of the mechanicals. So they had typesetters doing phototypesetting in another part of the building.

In the first sample we have an all-caps Futura headline in which the exclam-down is slightly shifted below the baseline... wheras in the U&lc sentence within the yellow word balloon, we again have Futura, but the exclam-down is aligned at the x-height.

In the second example the exclam-downs are lined up with the caps in an all-caps subhead, but in the third example the exclam-down is once again shifted slighty down, in spite of the all-caps headline...

My only guess is that in this case, in which all of the magazine's headlines were set in Futura, the variations in the vertical position of the exclam-down and question-down depended on the knowledge/care/obsessiveness of each individual typesetter at the plant... I never got to meet any of them, so I really don't know. But Kent's earlier observation certainly seems to apply here.

Ricardo Cordoba's picture

If nobody shows up I may send the URL around to see if we can get additional participation.

Eben, I can e-mail Eduardo Manso, and also Armin Vit -- I forgot to mention him earlier, and he is a frequenter of this forum.

ebensorkin's picture

The impression I get from emailing with Carl is that x height alignment is generally the trend at Monotype - existing designs excepted. Except in the case of all caps. They get Cap Height Aligned - Thanks Carl!

Ricardo Cordoba's picture

Hey, that's good to know -- so at least two big companies use the same standards. Thanks Eben and Carl!

Miguel Hernandez's picture

The dots hav eto be positioned on the top & on the bottom of the x height. A question in language have emphasis on the beginning and on the end on the question when is spoken, in english only at the end.

Miguel Hernández

Latinotype. The latin alphabet, on latino hands.
http://www.latinotype.com/view_fonts.html

Nick Shinn's picture

one uses small caps, but the questiondown still lines up with the small caps rather than the initial cap

That's because the small caps are faux.

Here's the breakdown:

1. Caps with small caps - wrong
2. Caps with small caps - correct
3. Caps with small caps - wrong
4. All small caps - correct
5. All caps - correct

Adobe Garamond Pro gives you (1) with both small caps, and all small caps settings.

Am I right about (2) Miguel -- or is (3) preferred? And if so, presumably because of the tails to cap Q and J?

Ricardo Cordoba's picture

Nick, I would say that in example (3), i.e., almost the same as the way the sentence appears in the book, the question marks are more easily seen by a reader than in example 4... The opening question mark in (4) is like a little squiggle, not a real question mark.

hrant's picture

I don't think the opening and closing marks need to be forced to be the same. Especially in a longer phrase I can picture a tall/high one to open and a "regular" one to close.

> presumably because of the tails to cap Q and J?

Indeed, this should be part of the -preferably dynamic/contextual-
determination of what kind of opening question mark looks best.

hhp

Ricardo Cordoba's picture

Not a bad idea... And with OpenType it is actually possible to do.

Miguel Hernandez's picture

I am with number 3 Nick.

Miguel Hernández

Latinotype. The latin alphabet, on latino hands.
http://www.latinotype.com/view_fonts.html

k.l.'s picture

These are very nice and instructive examples! I find the comic lettering particularly striking because it uses lowered exclamdowns despite the text is in all caps, and because handwritten, it shows that the positioning was chosen intentionally. In contrast, as Kent Lew says, it is not clear whether typographers decided for this or that positioning intentionally or just relied on the font's default.

Out of interest, since we touched kerning earlier: Are no quotation marks used in Spanish, so quotation mark to exclamdown/questiondown would be an unlikely combination?

I am most deligthed about the little spaces after ¡ and before ! in the title ¡YO NO FUI!

Nick Shinn's picture

it uses lowered exclamdowns despite the text is in all caps,

Lowered relative to what?! -- relative to the capitals that immediately follow, yes, but in relaton to the text as a whole, which wanders considerably from the baseline, no -- and that is logical, when one considers that the exclam down always prefigures by the final exclamup, and in so doing establishes a relationship with every character in the journey to the end of the sentence, and must harmonize with them all.

the little spaces after ¡ and before !

The exclams' spacing too should be contextual -- needed for an all-cap setting, and more so for sans than serif. Memo to OT developers: alternate exclam glyphs for "case" feature should have wider sidebearing.

Here's a typeface where there are no descenders in caps or small caps (notably, the Q and J).
Is option (3) still preferable to option (2)?

hrant's picture

> establishes a relationship with every character in the journey

Hmmm, very lyrical, but not exactly how people read. :-)

hhp

Nick Shinn's picture

not exactly how people read

No, but people also "read" the typography as a whole, and are aware of deviations and consistencies in the overall matrix. After all, it's against this background understanding of the implicit "big" grammatical structure -- which can extend to long sentences with many subordinate clauses -- that punctuation marks and page layouts have meaning. Emphases (parentheses, italics, bold) must also be read in this way, with a remembrance/anticipation of text outside immediate focus and attention.

hrant's picture

> are aware of deviations and consistencies in the overall matrix.

Not really.

hhp

ebensorkin's picture

Nick, thanks for expanding the Small Caps issue. You are right - it's more complicated.

Karsten said it shows that the positioning was chosen intentionally I feel slightly uneasy cleanly extrapolating from handmade comic lettering to text or even to display in type. That said, I think there is plenty to learn from handmade comic lettering, not least in this case. So I agree - mostly.

Miguel, thanks for checking in!

The thing that strikes me most about 1-5 is that it looks to my eye like it would be possible to design agreeable versions of both Cap aligned & x aligned question & exam marks. If I have time I will try to show what I mean.

The exclams’ spacing too should be contextual Of course I think it should all be contextual.... but that's a side issue.

Even though this thread has been interesting, I don't think my mind has been changed - just shifted. It seems valid to me to do either thing, but in general and for text especially I will favor Richard & Miguel's preference. As far as Nicks CAP examples. I think 3,4& 5 all look okay to me. 1 & 2 do not. But As I said, I think that's a function of the design of the faces used nearly as much as it is a function of what tends to look/feel best.

The ¡YO NO FUI! example is nice because It's a great example of where a Modern sensibility seems to be better served or at least is more consistent with the H align model.

Nick Shinn's picture

are aware of deviations and consistencies in the overall matrix.
Not really.

I've occasionally had the experience where the first first thing I notice about a page of type--the instant I look at it--is a typo. (or an f-i that isn't ligatured, or a straight quote mark). I wonder if other people have had the same experience.

How would you explain that?

hrant's picture

You're a type freak, not a target user.

I myself "notice" that lc "a"s are "backwards". But I do nothing about it.

hhp

Ricardo Cordoba's picture

The ¡YO NO FUI! example is nice because It’s a great example of where a Modern sensibility seems to be better served or at least is more consistent with the H align model.

Wanna hear something funny? Yesterday I was examining that book more closely, and on the spine the title is set in a different typeface, with a lowered exclamdown character! Talk about inconsistencies. :-)

I'll try to post a scan of it later.

Also, I am going to e-mail some colleagues in Argentina (two of them typography profs at the University of Buenos Aires), to see what they have to say about some of the questions that have come up in this thread.

Nick Shinn's picture

You’re a type freak, not a target user.

Sure, spotting typos is learned, but so are such simple activities as picking berries or lice eggs.

Not every text is as simple as a book page. Legal and government documents with their hierarchy of numbered paragraphs and sub-sections (and the Bible) are texts where page-context is important in reading. I'm guessing that not all saccadic scans follow the bouncing ball, and a brief internet search of scientific literature reveals that the "foraging" nature of saccadic eye movements has been documented.

The presence of an opening question/exclamation mark in some languages and not others suggests the complexity of the issue. How is it that English readers are able to start reading a long question with the final mark well in the distance? Are there sufficient grammatical cues in the opening words, or do we "catch" the question mark in "the corner of the eye" in a foraging saccade?

hrant's picture

> not all saccadic scans follow the bouncing ball

But the ones that don't are not immersive.

On the other hand I'm certainly not saying all
opening exclamation/question schemes need
to cater to immersive reading. Just that the
distinction has to be made.

hhp

Nick Shinn's picture

But the ones that don’t are not immersive.

Is immersion therefore primarily interpretive?
I'm imagining a concept rather like a "magnetic" Flash interaction, where the reader attempts to direct a concentrated flow of saccadic movements towards the next word, but foraging saccades are always bouncing off all over the place retrieving information that's mostly irrelevant, but nonetheless has to be processed in the feedback loop.

hrant's picture

> Is immersion therefore primarily interpretive?

I'm not sure what you mean.
What I'm getting at is that during immersion we don't care much
about consistency of alignment and other such conscious frills.

> foraging saccades

Errant saccades break immersion.

A good example here is when you start to read a question but only realize it's a question when you see the mark at the end, often resulting in a regression just to revisit what's been previously read. This is in fact why opening question marks are so useful. I wouldn't mind seeing them ported over to English! Or even better, do what Armenian does: allow question marks directly on the vowel(s) in question, so to speak. :-)

On the other hand, a reader can -and does- go in and out of immersion,
so there certainly can/should not be such a thing as a "pure" text face.

hhp

crossgrove's picture

"How is it that English readers are able to start reading a long question with the final mark well in the distance?"

Look at your own sentence. Spanish structure is less explicit and benefits from leading punctuation. English tends to put big honking clues at the beginning of sentences that are questions. Generally in English you invert the order of the statement and that signals a question. Spanish, not as much.

hrant's picture

> English tends to put big honking clues at
> the beginning of sentences that are questions.

Tends to.

And then there's bad English, which we still have to read and absorb!
"I asked Maria you wanted tuna or turkey sandwich?"

hhp

Syndicate content Syndicate content