WARNING: this question is misleading. See my answer below. Apologies and many thanks to drammock and Draconis for their answers too.

In general, for example in Times New Roman, U+027E has a fishhook at the bottom. But in Arial and several other fonts, it doesn't, as discussed on Wiktionary.

What should one understand from this? Could that character be defective in that many fonts? Or is the fishhook in LATIN SMALL LETTER R WITH FISHHOOK actually optional?

I highlight that the French name for U+027E is apparently LETTRE MINUSCULE LATINE R SANS OBIT, meaning it's named by what it lacks rather than what it adds.

  • 2
    I’m voting to close this question because it's about typography rather than linguistics.
    – Graham H.
    Commented Jan 11 at 18:43
  • No need to retract a question like that. It's a reasonable thing to ask (going back to errors in both Wikipedia and the Unicode standard) and that's what these sites are about: documenting knowledge for the future!
    – Draconis
    Commented Jan 13 at 6:44

3 Answers 3


@Draconis has given the core correct response, which is:

There aren't really any strict rules for what glyphs in a font have to look like—Unicode generally tries to encode semantics rather than appearance.


The name WITH FISHHOOK is somewhat misleading in this case

To that I'll add that I don't agree that the Times New Roman glyph for 027E actually has a fishhook at the bottom:

U+027E latin small letter R with fishhook

What you called "fishhook" is in fact just bilateral serifs, i.e. the sideways protrusions at the base of the stem of the lowercase r.

  • 1
    Agreed. There absolutely is not supposed to be a fishhook (which is not an actual Unicode diacritic, as far as I’m aware) at the bottom of ⟨ɾ⟩, or indeed any other kind of hook. The Unicode name is just completely bizarre. The lack of a head serif and fillet on the stem makes the top look more like a fish hook, but the bottom is just a bog-standard stem foot, with or without serifs as in the regular ⟨r⟩. Commented Jan 12 at 1:30
  • Thank you very much. That was indeed what I meant, incorrectly. I submitted an edit to your answer to... remove yours. Commented Jan 13 at 12:33

Unfortunately, I'm not sure how Times New Roman does it, since I don't have Windows and you haven't provided any images. But let's see what the Standard has to say about it:

image from the standard

excerpt from the standard

There aren't really any strict rules for what glyphs in a font have to look like—Unicode generally tries to encode semantics rather than appearance. But the reference glyphs in the chart don't have a hook at the bottom.

I've always taken "fishhook r" to mean it's an <r> shaped like a fishhook (that is, without a thingy at the top left), not an <r> plus a fishhook, and it seems that's how the Unicode Consortium interprets it too. The name WITH FISHHOOK is somewhat misleading in that case, but that's far from the only instance of a misleading name in the Standard. (Here's another example, off the top of my head.)


I am answering my own question after the answers from drammock and Draconis helped me realize that due to a mistake in Wikipedia, controversial usage of serifs on U+027E and U+027F by a minority of sans-serif fonts, including DejaVu Sans and LaserIPA's IPSsans, confusion from some resemblance with an anchor and inconsistent names and an English name which is far from clear, I confused the fishhook with serifs. What Unicode calls the fishhook is actually the "arm" (that name doesn't have consensus even for the regular r) at the top right, and what I called "fishhook" is actually the regular bottom with bilateral serifs.


It does make sense to call the r's arm a hook, since―even though it is not a diacritical hook―that is how the similarly shaped lowercase f's top is called, according to several sources (although these all have limited reliability and visibly largely copy each other):

  1. Typography Deconstructed
  2. Typography 101

A curved, protruding stroke in a terminal. Usually found on a lowercase f, it appears curved or bent like a hook.

  1. uxcel

The hook is the prominent curved stroke found in the terminal of the lowercase f, J, and j. As with some other typography terms, it was named after a real-life object it resembles — a fish hook

  1. TotallyType (Note that the part on hooks only shows after activating "Show Advanced Strokes")

The curved stroke in a terminal found in the f and r.

I find TotallyType's depiction most interesting since it shows variation in where the hook ends:hook on the f, according to TotallyType


The above definitions of hook can be confusing since they refer to "terminals", which seems to have various definitions:

  1. Google Fonts

the ending of the stroke 2. Typography Deconstructed the terminal is a type of curve. Many sources consider a terminal to be just the end (straight or curved) of any stroke that doesn’t include a serif (which can include serif fonts, such as the little stroke at the end of “n” as shown in the illustration). Some curved bits of tails, links, ears, and loops are considered terminals using the broader definition (see the Microsoft Typography site for further explanation).


The Unicode Standard 15.1 uses the term "fishhook" to name 7 glyphs, all based on lowercase letters:

Based on the Unicode Standard's depiction of U+027E and U+027F, I would have been inclined to conclude that what it calls a fishhook is a hook ending with a somewhat pointy "terminal" (not a ball terminal), making it more similar to a physical fishhook. But its depiction of U+02AE, U+02AF as well as U+1D73 disprove that hypothesis, as the following comparison shows:Unicode's depiction of U+027E, U+027F, U+02AE and U+02AF

One would think that U+1DF11 𝼑 would be the best example to infer what a fishhook is, but if so, it looks just like a regular hook with a teardrop terminal, as the normal "r" (U+0072) already has:

Unicode's depiction of U+1DF11

Most confusing when comparing with 5 other characters (U+1DF26 - 1DF2A) from the very same range (Latin Extended-G), whose hook is much sharper, yet always described as a simple hook: Unicode's depiction of U+1DF26 - 1DF2A


In conclusion, there is no defect in Arial or any of the other fonts I listed on Wiktionary. I reported the usage of serifs in DejaVu Sans 2.37 which contributed to my confusion, but DejaVu appears to be abandoned, so this will likely persist.

As Unicode visibly uses the term "fishhook" inconsistently or erroneously, its English name for U+027E (and others) should surely be changed to avoid that term, instead describing other specificities (as the French description of U+027E does), but Unicode character names cannot be changed anyway.

Thanks to Denis Moyogo Jacquerye for his insights

  • 1
    Unfortunately, once a name is in the Unicode standard it's there forever. They can add alternate names but never remove them.
    – Draconis
    Commented Jan 13 at 17:25
  • Thanks again Draconis; I edited my answer to reflect that. Commented Jan 15 at 0:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.