Hook styles on Ƙ Ƴ ƴ

Kent Lew

This is probably directed mostly to @Denis Moyogo Jacquerye, but I’m posting it publicly for the benefit of all and in case there are others with direct experience with West African language communities who might also have insights.

There appear to be two different prevailing approaches to the hook on the ƘƳƴ (and Ⱳⱳ) used in some African languages, especially in sans-serif designs.

One style is to have only a slight hook on the diagonal. This is often accompanied by a raised arm in the lowercase ƴ, and sometimes a little lift to the capital arms as well. (In a serif design, there will usually be a terminal treatment akin to that of lowercase c.)

The other style is to have a fully curved hook, usually at the cap- or x-height, though sometimes also raised in the lowercase. (In a serif design, the terminal may or may not echo other terminals as much.)

I’m interested in gaining any insights into the connotations or associations that the different styles might have for relevant language communities — e.g. Is one style more familiar or considered more readable than the other? Is one viewed as more formal/informal? Is one or the other more associated with or more appropriate for a particular design style or text vs. display use? etc.

Obviously, each presents its own challenges in terms of fitting and managing weight & stylistic harmony, etc. While deciding which to pursue, I’m interested to know what functional, cultural, or user-acceptance ramifications might also be factored in.

Igor Freiberger

I chose a closed hook to avoid confusion because an open form could be seem as an ornamental finish. Due to the lack of information, I included two heights for W and Y with hook. WIP:

Denis Moyogo Jacquerye

@Kent Lew Both forms seem to work well in your examples.

The fully curved form can be easier to distinguish at smaller sizes, but it’s true that it works better with serif styles.

Generally it may be helpful to see how salient the hooks are on other hooked letters Ɓ ɓ Ɗ ɗ Ƈ ƈ Ɗ ɗ Ɠ ɠ Ɦ ɦ Ƥ ƥ Ƭ ƭ or even on J j and others like Ŋ ŋ Ɲ ɲ. While we may be used to J having its hook extend below the baseline or fill a portion of the space left of the stem, we may not be used to see that with hooks above so it may feel odd. I don’t think we should shy away from salient hooks above, but they don’t always feel like the best option. Igor’s design seems to be a good compromise between the two.

One can also adjust the diagonal stem of Ƙ Ⱳ ⱳ Ƴ ƴ and start the curve earlier, for example like in Kigelia.

Kent Lew

@Denis Moyogo Jacquerye Thanks for your comments. Thanks also for the pointer to Kigelia. That design has the benefit of contrasted strokes, which relieves some of the pressure on the diagonal hooks, in terms of managing excess weight.

One of the design conceits of my project is that terminals are constrained to either vertical or horizontal endings. This tends to put limits the room to maneuver in the compromise space in-between.

Since you asked, here is the range of hook treatments:

The open diagonal hooks reference the ascender hooks; whereas the fully curved treatment echoes the horizontal hooks of ƁƊƤ et al.

As you accurately point out: the more salient treatment of hooks is further outside my western Latin comfort zone. My initial inclination was the more open treatment, which harmonizes more obviously in my English-accustomed eyes.

But I have grown concerned about the subtlety of the uppercase ƘⱲƳ in running text proofs. (Less so with the lowercase ⱳƴ, since the raised arm signals quite well.)

FWIW, I did explore a “hard” treatment of fully curved hooks, drawing more explicitly on the horizontal capital hooks, which mitigates the weakness at the top of the arms, especially in the heavier weights:

But I had rejected this direction as probably too unconventional and disjointed.

However, you’ve confirmed my instinct to continue striving to find a workable solution for more salient hooks (at least in the case of the capitals). So thanks for that.

Kent Lew

(It’s really disconcerting that the Times New Roman fallback font that I see displays abjectly incorrect forms for Ƙ and Ƴ. 😖)

Denis Moyogo Jacquerye

Oddly enough the metal Times New Roman had the adequate form for Ƙ:

 

The mirrored Ƴ was used in the fonts used in the Unicode charts up to version 5.0 ( see https://www.unicode.org/L2/L2004/04399-y-hook-defect-02n3768.pdf ).

John Hudson

I recall someone at Microsoft querying my design of hooked-Y with the hook on the right, on the basis that it was different from TNR and the Unicode chart.

[Are you able to edit that link to remove the closing parenthesis?]

PabloImpallari

On LibreBaskerville African extension I derived most of the hooks using the /j as the initial reference form, rotated and tweaked according to each letter.
However for the /Ƙ /Ƴ /ƴ hooks I used /ɂ as the reference, which in turn uses the /questionmark as reference.

Igor Freiberger

@John Hudson
There was an old discussion on SIL forums about the left x right position of the Y-hook. Linguists who wrote there pointed they know about a possible left hook, but found no evidences of it in their contact with Niger-Cameroon languages. I got no further information about it so I relay on the right one, which is confirmed.

Gentium includes both styles in uppercase, but only the right one for lowercase:

John Hudson

In Brill, I made the hooks quite reserved, relying more on weight than size to distinguish from regular Y/y, and trying to incorporate them more naturally into the texture, and reflecing the handling of terminals in the Cyrillic and elsewhere in the design. I had some leeway in this, since Brill were not working extensively with languages that use these letters, but it is the sort of decision I query them about when we come to do updates.

Kent Lew

@Denis Moyogo Jacquerye How unorthodox would it seem to a native reader to have full curved hooks in the capital Ƴ paired with shallow open hook on the lowercase ƴ ?

Would this pass largely unnoticed? As noted previously, the fuller hooks would provide more differentiation, and thus presumably readability, in text sizes.

Although the shallow open hooks provide for more even spacing, especially in larger sizes.

Igor Freiberger

@Kent Lew
Although I use closed hooks in my own fonts, for your font I prefer the open solution. In the samples above, the second (with no mixed hooks) seems definitively better to me. You could achieve more 'hooky' hooks with a 45° cut, but I believe the cuts at 90° are a design criterion you prefer to keep.

Kent Lew

From a design perspective, I agree with you. But from a reading perspective, I’m not so sure. As a non-reader, I can’t really judge well whether the open hooks read adequately enough or not.

For example, in these snippets from Hausa Wikipedia, compare the hooked Ƙanƙara at the beginning of several lines with the plain Kudancin at the beginning of the final text line.

My gut tells me that the extra expression of the full hooks is beneficial. But perhaps that’s not necessary for a native reader. Without the benefit of semantic discrimination, it’s hard for me to say.

You’re undoubtedly right that introducing a non-orthogonal termination would permit more of a compromise. Call me obstinate, but that’s not an exception to the design constraint that I’m prepared to make yet.

John Hudson

Maybe make the angle of the arm steeper, so you can accomodate a longer hook? You could try to create the same amount of space on the underside of the hook as you are able to get in the lowercase form. The problem is not necessarily or only the shape of the hook, how prominent the feature is.

Kent Lew

Interesting suggestion, John. Will investigate. Thanks.

Craig Eliason

I'm picturing a "Linotype f" treatment...