TeX - LaTeX Stack Exchange is a question and answer site for users of TeX, LaTeX, ConTeXt, and related typesetting systems. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I often work with texts from the 17- to 1800s that (in addition to now obsolete words) contain familiar words in slightly different spelling. The language of these texts, and of the ones I'm writing, is one in which a lot of hyphenation has to be done, due to the high average length of its words (German). To complicate matters further, it's also a language in which hyphenation is regulated a lot more strictly than in, say, English. [Which doesn't mean that what follows is a problem restricted to German.]

In *TeX, this results in a problem: when hyphenating a word that comes in an archaic spelling, *TeX has no choice but to apply the same rules it applies everywhere else, as it has no way of telling if the spelling at hand is archaic or not. Quite often, this results in a hyphenation that is wrong by both archaic and contemporary German standards.

A classic example:

A lot of German words used to be spelled with a th where today there's only a t. Such as Teil (=part), which was Theil until that th went out of fashion, and even became eliminated from the official dictionaries around 1900 -- probably because of it's redundancy, as the h didn't make a difference in sound, unlike in English. Today, only few German words retain that Th, such as Thron (=throne), as they didn't want to offend the emperor when they officially changed the spelling rules.

The reason this is important to *TeX is that in today's German, with that kind of th eliminated, the letters t and h hardly ever constitute one sound. So if we encounter a modern German word with a th in it, which happens quite often, it will usually be a word composed of two or more syllables (or words) -- the first of which ending with a t, the second starting with an h. As in, say, flathead in English. But sound is the very basis of German hyphenation practice, and *TeX knows that. So in *TeX, the archaic spelling of theil in beurtheilen (=to judge), will result in the hyphenation be-urt-heil-en, which is unacceptable because, in this case, the t and the h belong together in terms of both sound and meaning.

My questions are: has this ever been addressed somehow? I haven't been able to find anything suggesting that is has. Do we have other options except manual word lists using \hyphenation{...}? ...Which is particularly tedious in German because every potentially critical word may come with a dozen of versions for different genders, cases, etc.

share|improve this question
Although I have no answer or knowledge of German, I found this question informative and interesting. – Ryan Reich Mar 19 '13 at 15:28
I don't know if there is a babel package to take care of this, but I guess not or you would have found it. I have used hyphenation{...} for texts with lots of foreign words. I didn't put any words in there until I was finished writing the whole thing and actually saw where the problems occurred. If you write these words very often, you could just have your own little hyphenation "library", kind of like your .bib file that grows over time. – Anke Mar 19 '13 at 15:35
That's exactly what I've been preparing for, mentally, but of course I want to make sure if there's an alternative. And I'm writing a mere PhD thesis -- I recently discovered that an 8-volume edition of an 18th-century author I've been dealing with was typeset in *TeX. If there's really no way around \hyphenation{...}, I don't envy the people who did the typesetting. I might try to contact the publisher and get in touch with those people just to ask them about this... – Nils L Mar 19 '13 at 15:41
up vote 7 down vote accepted

The canonical approach is to use dedicated hyphenation patterns. I'm not aware of existing patterns for German spelling predating 1901, but with a little bit of work, it should be possible to prepare some.

As part of Google's Ngram Viewer you can find corpora of words with annotated publication years for different languages (among them German) that could be used to prepare a list of German words from books published between, say, 1800 and 1900. Such a list needs to be hyphenated (can be done with existing German patterns) and checked for the wrong cases you described. Finally, patterns can be generated from that list.

I'd like to invite you to join German pattern project at trennmuster@dante.de. We can discuss the issue in more detail over there.

share|improve this answer
yes, I remember reading about that in DTK a while ago. Thanks for reminding me, I'll get touch with you guys... – Nils L Mar 24 '13 at 17:40

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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