I use ligatures. They are a beautiful typographic tradition and many fonts support them. Unfortunately the rules to apply ligatures are not hard and fast.

One rule says that you use ligatures only inside a morpheme [https://en.wikipedia.org/wiki/Morpheme], which seems to be a pretty difficult subject.

Another rule (of thumb) says that you do not use ligatures across syllables.

But what do you suggest, if these two rules conflict?

I came across a single example, which is in German, a language which is more prone to misplaced ligatures, as German has many words which are composed of other words, but in those cases the rules are easily applied.

Would you use a ligature in the German word 'Hüfte' [hip]?

The stem is 'Hüft-', that would mean: yes.

But the syllables are 'Hüf|te', which might say: no.

I know that this is not a TeX question per se, but I know that many TeXnicians are also typography aficinoados, that's why I am hoping for some good advice here. Am I overthinking this?

  • 5
    I only know of the rule that you should not use ligatures that cross word joints (Wortfugen), which indeed are quite common in German. Thus, in a word such as "Kaufleute" (which consists of "Kauf" and "Leute") no ligature is used, but in "Kaufläche" (consisting of "Kau" and "Fläche") you would use a ligature (examples taken from Wikipedia). Your example "Hüfte" does not include a word joint and "Hüfte" consists of only one morphemes (I think). The Duden gives as example for the use of ligatures the word "heftig" where the ligature crosses two syllables. So, I would use a ligature in "Hüfte". Feb 23, 2021 at 10:40
  • 1
    Your example with 'heftig' convinced me that the syllables rule is probably wrong. I'll mark this as 'answered'. Feb 23, 2021 at 11:18
  • 2
    Are you familiar with the selnolig package?
    – Mico
    Feb 23, 2021 at 11:19
  • 2
    @StefanSchroeder - Why do you suggest that the package's rules are "based on heuristics"? The package's German language selective ligature suppression rules were based on a list of several hundred thousand words analyzed by a morpheme lexical parser. (Full disclosure: I'm the author of the package.)
    – Mico
    Feb 23, 2021 at 11:27
  • 2
    And an excellent package it is (speaking from experience) :-)
    – Ingmar
    Feb 23, 2021 at 11:27


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.