I'm curious if there is a form of the AE diphthong that is preferred.
The one on the left is the more standard, the one on the right is easier to space with other glyphs.
The one on the right seems an A with speed lines! ;-)
Jokes aside, the problem is that in it the E is not recognizable as an E. Or, at most, it appears backslanted.
I've seen both used. Is there a preference?
I'm not familiar with the right one version. I think the best reference would be some Roman inscription.
But maybe some Danish or Norwegian could chime in and shed some light from the perspective of a modern user.
I don't believe classical Latin used Æ - it's a medieval invention. I only find blackletter examples.
I think the best reference would be some Roman inscription.
… if only those old guys could have made up their mind!Thomaskirche Leipzig, 1593
I don't believe classical Latin used Æ - it's a medieval invention.
Not so. From what I gather, It was often used in inscriptions. But mostly at the end of words and essentially just in order to save space.
Incidentally, the samples I was able to find were of the "backslanted E" type.
… if only those old guys could have made up their mind! Thomaskirche Leipzig, 1593
Well, I meant Latin as in Roman Empire, but that’s a nice example of OpenType contextual alternate! ;-)
From what I gather, It was often used in inscriptions. But mostly at the end of words and essentially just in order to save space.
Latin inscriptions often contained not just Æ but other space-saving ligatures. So Æ was essentially just a graphic variant of AE and not accorded special status in Classical Latin. But then the diphthong represented by the digraph AE began to be pronounced as a single sound in Later Latin, eventually merging with E, and the ligature Æ became a useful shorthand for this.
Want a kind of logical answer?
If in the lower case, the e is dominant over the a (what I mean is, it retains its shape at the expense of the a whose tail is markedly affected by the e) then shouldn't the E similarly dominate the A (as it does in the vast majority of AEs) and thus the backbone of the E will be vertical?
I personally prefer this dominant E and don't warm at all to the slanted E but I guess you didn't ask about either preference or logic.
Bliss and Myriad use slightly slanted backbone to E. I guess If you're going for equal rights then you've got to slant it a little.
It was almost certainly you (or maybe Chris L) who once said to me, "whatever looks best is the right answer." Would this be any different?
Us Norwegians actually use it on a daily basis. Your first Æ is the only one I would ever consider using for text.
That said, you can always experiment with something in between.
My Latin is a little rusty.
Are there words where the sequence a_e may not be "diphthongized"?
The left one, no doubt. In Norwegian and Danish, the E part is the important one. In a lot of words, it's pronounced like e, while it's never pronounced like a. In Icelandic, however, it's pronounced almost like the old Latin diphthong. Their preferences may be different.
Interesting. I thought this (and OE, and oe, and ae) were a matter of getting balanced shapes to fit together. It hadn't occurred to me that dominance of one of the letters would be intentional, and linguistic rather than visual.
Gee, I have absolutely no authority behind me on this (besides being Norwegian), this is just how I feel about the matter.
Are there words where the sequence a_e may not be "diphthongized"?
Not as far as I remember (not that my latin isn't rusted too).
Let's compare the design possibilities, also taking the related ligature Œ into account.
1. E is the dominant element. This seems to be the most common solution by far.
2. Both elements are equally dominant. I don't really recall seeing this type of solution, even though it doesn't look entirely displeasing to the eye.
3. The A or O is the dominant element. Rarer than 1. but does occur.
4. The elements retain their own shapes and are joined at a point. Much more common to see the Œ designed this way than the Æ.
5. Erm, yeah. Overlapping elements. Train wreck for the Æ, but for certain typefaces the Œ is designed this way.
Interestingly, the Æ is likely to take form 1. even if the Œ takes other forms.
Nick: Are there words where the sequence a_e may not be "diphthongized"?
In ecclesiastical Latin one encounters words of Hebrew origin, usually names, in which ae may be pronounced as separate vowels. In this case, though, the variant pronunciation is typically indicated by a diaeresis, e.g. Sancte Michaël Archangeli.
I always use (1) for AE, and for OE in most weights. In bolder weights, (4) works well for OE.
If it doesn't look too obvious, I try to make these characters narrower than they would be merely by overlapping their constituent parts.
Script fonts with swash capitals are quite a challenge for these characters.
Don't forget to kern P and Þ with Æ.
Is there any merit in aligning the crossbars in Æ?
My Latin is a little rusty. Are there words where the sequence a_e may not be "diphthongized"?
Yes, there are some loanwords and proper nouns, like Aetius, where "ae" represents two separate sounds rather than a diphthong. In Roman inscriptions, though, whether it is a diphthong has nothing to do with whether it is ligatured. Any pair (or occasionally more) of letters in any word can be ligatured, if it makes sense graphically. It's purely a design/layout decision.
(ETA - I should check for new posts before hitting the button...)
The problem with Æ no. 2 is the angle of the stem. Since it's neither vertical nor really diagonal, it makes the glyph look left-leaning ("linkskursiv" or "rückwärts liegend" in German, I don't know the English term, if there is one). My choices are definitely Æ no. 1 and Œ no. 3, but for a sans serif I might chose Œ no. 1. Œ no. 4 is pleasing to the eye, but I feel it's not ligated enough. Œ no. 5 looks very Mannerist or Baroque, and might work in such a typeface.
Just as I think the E is the important part of the Æ ligature, I think the O bit is most important for the Œ. I'm of course heavily culturally biased. The Œ is not used in modern Norwegian, but it's traditionally used for normalised transcription of the long Ø-sound in Old Norse. (These days, it's sadly being replaced by the (mostly) inaccessible and ugly Ø acute. Curse modern scholars!)
I almost always go for a solution like on the left of the initial post - but this is just my personal preference. This also has has the added benefit of the left side kerning like an A and the right side kerning like an E for the class kerning :)
I also prefer the Æ to be narrower than its two parts - so shortening the E bars and closing up the counters of the A bit is par for the course.
But be aware that this Æ form is sort of constructed. In written form the Æ is written in just three strokes:
- the A angle and lower bar of E in one stroke: /\_
- the crossbar of the A and E in one stroke: --
- the top bar of the E starting at the apex of the A:-
So the right solution of the initial post is closer to a written form.
@ Nick Shinn: yes, the crossbars should ideally be aligned - but this is usually difficult to achieve in bolder weights (the A counter becomes too small very fast ;)
There are many horribly designed "Æ"s out there, so be careful what you use for reference...
the crossbars should ideally be aligned
I disagree. That depends on the typeface. For most serifs, I'd say it more often than not looks forced and disproportional.
@ Satyagraha: I said "ideally" not "preferably" ;)
Usually it is not possible or a viable option to design the Æ with a common crossbar - my point was just that it is closer to the written form...
I have never drawn an Æ with aligned crossbars, but I do in handwriting.
...whether it is a diphthong has nothing to do with whether it is ligatured...
That's what I thought.
I put a "Latin dipthong" feature in Paradigm Pro as a Stylistic Set, including "titlecase" glyphs.
Bit of a gimmick, and I really wanted to include roman numeralization too, but never got around to figuring out the code (although it was published here recently).
"It was almost certainly you (or maybe Chris L) who once said to me, "whatever looks best is the right answer." Would this be any different?"
As long as you factor in what native speakers of affected languages read properly. This matters most for text fonts--display faces allow a bit more leeway. Ask the Scandinavians what "looks best" to them.
For what it's worth, I find Œ no. 1 most natural for French. Granted, it's not the most commonly occurring letter, since so few words begin with it. I think Œ no. 4 works best when the capital letters are quite wide and old-fashioned. Œ no. 5 I can't really stand.
I would also align the crossbar of Æ in handwriting, but the letter is not commonly used in the languages I speak. I remember seeing a design where the crossbar is a single slanted stroke that rises to the right, but I can't find it.
No AE or OE combinations to speak of, but one of my favorite "space saving" inscriptional arrangements. From the Arsenale in Venice:
That Benguiatus was a good carver!
I think this almost works, but it's still a slightly reverse cursive Æ to my eyes. Notice the "cheating" with the stem and the counter necessary to achieve balance. (I'd forgotten what a great typeface Warnock is.)
Definately the left one. Think of it like an E with an A attached rather than the other way around. Like Frode I would ever use the one on the right.
Warnock's Æ, compared with A and E:
Sans serif solutions.
This is a monospaced font I am preparing (so outlines are still rough):
OE (and per thousand are a real headache) but I might go with the stacking OE and have a 'Natural Width' as an alternate/SS at double width?
Michael, I don't think your stacked OE reads as that letter, nor is it a necessary contortion. You've already demonstrated that you can make AE work on this width, and OE should be easier.
This is the very reasonable construction from Lucas' Consolas monospaced font:
Seeing Michael's example reminded me vaguely of KLTF Grotext. It has AE, OE, and UE (!) ligatures. In addition to the usual capital ligatures, there are glyphs where the first component is a capital and the E is in small caps. There are also Ae, Oe, and Ue ligatures where the second component is in lowercase form.
My guess is that these were conceived as experimental alternate forms of the German umlauted vowels (hence the UE ligature for Ü). These forms are certainly not what readers would expect for languages whose orthographies actually use æ or œ.
...these forms are certainly not what readers would expect for languages whose orthographies actually use æ or œ.
I don't think that's the case with œ in French.
In Canada, one is quite used to seeing œ spelled oeuf.
I don't know if it's the same in France.
At any rate, designers don't seem to mind splitting the diphthong, for various reasons:
I don't think that's the case with œ in French.
In France, too, œ is very frequently spelled oe, even though it is not strictly correct, in large part because it can't be typed with the standard French keyboard. I was talking about the actual ligature for text faces; when people go through the trouble of using the actual ligature rather than Oe, they would use the form Œ.
It is such a rare occurrence for Œ to come at the beginning of a word where only the first letter would be capitalized that I guess you can't really talk with much certainty about what readers would expect. In fact, I have to say the Oe ligature reads very naturally. I think French readers may think of œ as something like just tightly kerned oe, so the Oe ligature makes sense in that way. This would not be the case with, say, Norwegians and æ.
Here's another example I found, presumably from France (unless it's just a coincidence through bad spacing).
(Note that this is an exception. Most settings of Œdipe that I found through an image search use the standard form of the ligature or spell it Oedipe.)