Hans_Adler's messages on the Wall (total 8)

2016-02-28 10:43
I saw your comment over at [#783827]. For me this was helpful in one instance (where I had been in error about the spelling reform) and just plain wrong in another (your spell checker misinterpreted a legitimate past tense verb form of one verb as a mis-spelled present tense form of another).

I see many legitimate typos on your list, but also numerous instances of "ss" for "ß". The latter are legitimate Swiss German variant spellings. While technically regional rather than wrong, I think we should not accept this variant spelling as it loses information and can confuse non-native speakers of German.
2011-06-07 22:58
In languages with vowel harmony (e.g. Finnish, Estonian, Hungarian, Turkish) any inflections of a name will usually depend on the quality of the vowels in the name. I am not sufficiently familiar with any of these languages, but I would be surprised if it were impossible to build plurals or accusatives of names in all of them. The precise choice of name can also have a subtle influence on translations, as it may imply a setting (e.g. Tom vs. Julius Caesar vs. Ossama bin Laden).
2011-03-20 14:10
Was ich eigentlich noch sagen wollte: Sobald man ohne ersichtlichen Grund eine weniger übliche Übersetzung wählt (also etwa "Nutzer" statt "Benutzer"), neigen die Leser dazu, eher im Sinne der Grundbedeutungen zu denken. Bei "Computernutzer" oder "Bibliotheksnutzer" wäre das kein Problem, bei "Vertrauensnutzer" oder "Kindernutzer"/"Kindnutzer" (für "child user") dagegen sehr wohl.
2011-03-20 14:06
Dagegen spricht aber, dass im Computerkontext die Standardübersetzung von "user" "Benutzer" ist (oder Anwender). Siehe z.B. Google-Suche nach "Benutzerkonto" und "Nutzerkonto" (jeweils mit Anführungszeichen). Übrigens -- nicht, dass das ein zwingendes Argument wäre, aber die EU-Bürokratie hält es auch so: , en -> de einstellen und nach "user" suchen.
2011-03-20 06:20
Wenn schon "Vertrauens-X", dann auf jeden Fall "Benutzer", nicht "Nutzer", denn ein Nutzer ist kein User sondern jemand, der etwas benutzt.

Das Problem mit "vertrauenswürdig" sehe ich auch. Wenn es das deshalb partout nicht sein soll, sehe ich keine Alternative zum freieren Umschreiben. Mir ist dazu "etablierter Benutzer" eingefallen, und laut Google scheinen das auch schon einige wenige deutschsprachige Webforen zu verwenden. (Sehr viele lassen es einfach bei "Trusted User", was für uns wohl nicht so angemessen ist.)
2011-03-08 10:33
I doubt it would slow things down much, and would guess that the real problem is one of implementation priorities. But if there really is a performance problem, it could be solved by doing the sorting on the client side using JavaScript.
2011-03-08 10:30
Ooops. I just noticed that I seriously claimed that "Schmidt" could be read as an occupation. It appears when I am writing in English I cease to be a native speaker of German... Of course there are other examples that actually work, such as "Schuster".
2011-03-07 19:00
I think if we had a clear use case in mind translating names might be reasonable -- depending on what that use case is. However, as a general-purpose project that will probably be most useful in conjunction with automated procedures such as the one behind Google translations, we should avoid making things harder than necessary.

It's hard for a computer to decide whether a "Schmidt" in a German text (no preceding "Herr") refers to someone with that name (=> no translation) or to someone with that occupation (=> translation). Done right, our corpus will assist with learning the distinction. Done wrong, it will add to the confusion.

It's also not clear where to stop. Should we "translate" the many occurrences of Tokyo in our corpus by Paris, Berlin, Madrid etc.? I don't think so.

Related problems that already exist in the real world are described in this Language Log post: (In certain situations Google translated "Austria" as "Ireland" when translating from German to English.)