#### Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

# Dutch IJ with dots

• Posts: 192
If the design of the "f" and the "i" in a particular typeface is such that an "fi" ligature is needed, then, given that the design of an "ij" will be similar to that of the individual letters "i" and "j" in that typeface, it's hard for me to see how ligation of a preceding f to the following ij can be avoided.

Yet, Hrant's point is valid: since "ij" is a single letter, having an f preceding an ij have the action of converting it into an "fi" followed by a "j" is simply wrong.

If the letter "ij" can be distinguished from the individual letters i followed by j, though, the obvious solution is to define an "fij" ligature. As long as the "j" is shown to be linked to the i by the existing mechanism in the typeface, that the "f" is even more strongly linked by actually being connected to it... is not too bad.

But I'll admit that it's still problematic. Given the relationship between ij and y, it is definitely possible that the obvious solution - extending the line from the f that joins to the serif on the i to also join to the serif on the j - may look strange to native speakers of Dutch, and not accord well with their established reading habits.

That is, if the i and j in ij are to be linked, it seems that they're supposed to be linked on the bottom, not the top.

If that is the case, one could do the following - although I forbear to actually recommend this possibility that I'm pointing out - when ij follows an f, convert it to a y-umlaut, and then ligate the f to that in order to avoid the dot collision problem!
• Posts: 204
@Paul van der Laan The best test would be to take those three fonts and put a somewhat wide (and probably slightly taller) acute* [visually] centered between "I" and "J" (and "i" and "j"). One interesting resultant question: could the tittles provide an avenue of unifying a [non-accented] "ij"? Not least to avoid the "y" structure.
Visualize it. Take the fonts and show the world what you mean.
• Posts: 1,105
edited November 2017
People who rely on IJ much more than me should be eager to explore its potential. But that's assuming they're open-minded. In any case, here you go:

Do the single-acute ones look better to me only because I'm not used to the double-acute form? Quite possibly, but a much better question is: Over time would it end up looking –and most of all, working– better even to those who might be used to the double-acute form? I think so.

If you do want two diacritics, better than two "agnostic" acutes would be some variation that aids the unity of the letter, especially when the "I" and "J" are distinct like in that serif.

As an aside, concerning Plex: http://typedrawers.com/discussion/2422/plex-ibms-new-font-identity-model
• Posts: 1,119
The single-acute versions look extremely weird to me.
• Posts: 26
Thanks Hrant.

Double acutes are okay-ish, because I'm used to them.
The single acute on the ij-glyph (top line) is viable, and I like it, but this one is a matter of taste.
Singe acute on i-j combination (bottom line) looks like a mistake or error to me...
• Posts: 230

I am not used to see accents on uppercase letters in Dutch. Because of this, I am afraid, for me, all the above acutes on an uppercase “IJ”, look weird. (Perhaps it will look less weird in small text.)

• Posts: 1,105
edited November 2017
Ben Blom said:

To me, the whole exercise of adding acutes to an uppercase “IJ”, is of very limited use. The chance that anybody would add acutes to an uppercase “IJ”, is very very small.

You should see what we kern...

@Laurenz van Gaalen Thanks for the interesting feedback. On the one hand a single acute does look more natural on an already-unified glyph; on the other hand a non-unified glyph needs it more...
• Posts: 192
edited November 2017
Here is an image showing the possible alternatives I can imagine for a ligature of f with a subsequent ij:

In addition to the extreme approach of converting ij to a y-umlaut, a fourth choice, using u-umlaut as the basis is shown, based on the ij in some infanta sans-serifs shown in this thread.

Come to think of it, however,

a stealth approach is also possible - now the i and j are joined from below, but just by lengthening a serif on the i rather than by making a major visual alteration to the ij.
• Posts: 192
There is yet another possibility:

Given that normally the i and j are made into a unit simply by kerning them more tightly than i followed by j as two separate letters, if the issue is that ligating the f together with the leading i makes the f and the i as a unit, breaking away the j, then this could also be counteracted by increasing the physical proximity between the i and the j, which is already what is being used to make them a unit.
• Posts: 1,105
Given that normally the i and j are made into a unit simply by kerning them more tightly
Except touching is whole other level of tightness...
• Posts: 230
edited November 2017

John, one of the possible alternatives for a ligature of “f” with a subsequent “ij”, is missing from your examples. It is, in fact, just a “fi” ligature of the unconnected type. The first sample word below is without ligature; the second sample word has a “fi” ligature. (The font is Aspira. No complicated “eye salad” to look at.)

• Posts: 1,105
edited November 2017
If Chinese isn't too complicated, a meaning unification of "i" and "j" certainly is not. In contrast, Morse Code is supremely simple, but tastes like this: http://www.bbc.com/news/world-latin-america-41698885
• Posts: 1,105
* a meaningful unification
• Posts: 192
Ben Blom said:

John, one of the possible alternatives for a ligature of “f” with a subsequent “ij”, is missing from your examples. It is, in fact, just a “fi” ligature of the unconnected type.

Yes, I was only pointing out alternatives to the infamous "pot-hook f". But here we are for completness:

in the kind of Roman typeface that I was thinking would be more likely to have a problem. You are right that a sans-serif typeface would probably be able to go with the simplest alternative.

So, from left to right, first we have the pot-hook f, followed by both the current "ij" and one more closely spaced, then a conventional "fi" ligature combined with those two kinds of "ij", followed by ligating the f to an i and j that are also physically joined in four possible ways.
• Posts: 192
Here's some more options.
Those options seem... whimsical... to me, and thus to be unlikely to be essentially transparent within the existing reading habits of native speakers of Dutch.

Not that there's anything wrong with lightening up the thread.

And I'm well aware that some of the more pedestrian alternatives that I proposed likely wouldn't be acceptable to Dutch readers either. It's up to them to decide what works for them; but I'm hoping that a workable way of ligating the f while retaining the identity of "ij" as a letter is included among those alternatives.
• Posts: 230
edited November 2017
Artur, your examples may be more artsy—but, I’m afraid, a little too silly. To your second example, my response would be: Why this desire to connect things? The advantage of an unconnected “fi” ligature in the current context, is that it makes “f” and “i” visually look less connected, instead of more connected. This helps readers to see “ij” and “ie” more as “one thing”.

So, whenever the design of a font would allow it, an unconnected “fi” ligature would be the best way to ligate the “f” while retaining the identity of “ij” as a letter.

• Posts: 44
edited November 2017
Ben Blom said:
Artur, your examples may be more artsy—but, I’m afraid, a little too silly.
Yes ofcourse they are. As John noticed, just lightening up the thread as I feel ij is being turned into a problem, where from a native speaker point of view there is actually not much of a problem. We can debate about ij being viewed as one or two letters, about ij needing two acutes for emphasis, about a ligature or not, but fact is that ij never causes miscommunication. Only when some engraver suddenly feels the need to place two tittles on IJ is when things start to get confusing.

As far as I'm concerned, everything is koek en ei.
• Posts: 26
koek en ei it is
• Posts: 216
I finalised the font with two accents
https://www.behance.net/gallery/58756601/Combax-font
• Posts: 44
Artur Schmal said:
‘...but fact is that ij never causes miscommunication.’
Oh well, I'm reading now that the story goes that the father of Johan Cruijff had the name CRUYFF in his passport, but his mother used CRUIJFF.
• Posts: 1,105
edited November 2017
Artur Schmal said:
where from a native speaker point of view there is actually not much of a problem.
Before the invention of shoes people thought going barefoot was fine.
Hopefully others will take your own playful speculation as a mind-opener.

And details matter. Like the missing cedilla for a woman I once knew by the name of Anush being called over the Istanbul airport PA system... (True story.)
• Posts: 192
edited November 2017
I feel ij is being turned into a problem, where from a native speaker point of view there is actually not much of a problem.
Of course there's a problem. The problem is if you're not a native speaker of Dutch, you're designing a new typeface for which you want to support at least the major Latin-script languages, and it happens to need the fi ligature.

If the issue that Hrant raised is correct, that ligating f to the i in ij does something bad, then the lack of a recognized convention for addressing that case would be a problem. (The poor fellow doesn't know what to do, without making a significant effort in research.)

Of course, from this thread, it appears that most native Dutch speakers share the view I had before I heard of "ij" as a letter in the Dutch alphabet - it's a common digraph in their language, but either that's all there is to it or at least the letter "ij" is not considered a particularly important sacred piece of their cultural distinctiveness.

Given all the nationalistic things people do make a fuss over, and the often negative consequences of same, despite the other side of the coin, with examples like the loss of such things as indigenous languages, which is highly regrettable, I can't get overly agitated by this particular instance of apathy.

But if the Dutch change their minds, so this becomes an issue, well, they have plenty of options. Even ones as outré as letting Dutch be Afrikaans.

• Posts: 1,105
Culture is nationalistic.
• Posts: 739
“Do people in cursive handwriting also have an issue with cursive connections in these same scenarios?”

...that’s what I’d like to know too. Reading, writing, speaking, and hearing are the informants, typography is a disformant overlay.
• Posts: 230
edited November 2017

Traditional Dutch cursive handwriting as taught at school, would look like this:

This is a comparison between the traditional connected handwriting, and the alternative unconnected handwriting (which is taught at some schools):

• Posts: 1,105
edited November 2017
@John Hudson I will never forget my erstwhile calligraphy instructor's maxim: "The highest compliment is when your work is mistaken for a font."  :-/

@Ben Blom:

• Posts: 192