Left- and right-dislocation in German behave differently regarding the case the dislocated expression takes. Left-dislocation seems to be lenient, as it allows the nominative as well as the case the resumptive pronoun takes:

 (1) a. DER MANN, DEN kenne ich nicht.
     b. DEN MANN, DEN kenne ich nicht.
        'The man, I don't know him.'

In (1a), the dislocated expression DER MANN appears in the nominative case, even though the resumptive pronoun DEN is in the accusative case. In (1b), the dislocated expression is in the accusative, thereby matching the case of its resumptive pronoun. I am German native speaker, and for me both (1a) and (1b) are good.
Compare these data to those of right-dislocation:

 (2) a. * DEN kenne ich nicht, DER MANN.
     b.   DEN kenne ich nicht, DEN MANN.
          'I don't know him, the man.'

Example (2a), where the right-dislocated expression appears in the nominative is now ungrammatical. If the case of the right-dislocated expression matches, as in (2b), the sentence is good.

What principle could account for this asymmetry?

1 Answer 1


I think these sentences are structurally different. In (1a) “der Mann” is what we used to call “nominativus pendens”, but which nowadays is usually called “topic”. For inflected languages like German, Greek, Latin, the old terminology has the advantage that it points out that the topic is in the unmarked (i.e., nominative) form, and not in the case required by the syntax of the main clause. In German, and I think in most languages, the topic needs to be announced at the beginning of the sentence. That is why (2a) does not work.

In (1b) and (2b) the direct object (“den Mann”) has been extrapositioned, and its normal place in the sentence has been filled by a pronoun. This, I suggest, is a different structure.

  • +1 for pointing out that 1a and 1b are structurally different. I would add that (for me as a native speaker) 1a is only grammatical, if DER MANN is produced in a separate intonation phrase (in difference to 1b, where a single intonation phrase suffices for the whole sentence).
    – robert
    Sep 4, 2014 at 14:47
  • 1
    What does the structual difference across (1a) and (1b) look like in the syntax? Would one attempt to capture it in the tree structure? Sep 5, 2014 at 0:24
  • I, too, want to know more about structural difference. Can you provide sources or representations? Sep 5, 2014 at 14:05
  • @robert In my German (I'm a native speaker, too), both intonation contours are possible: one phrase, or two; rising melody as in questions is possible in both cases, but not necessary. Concerning this matter, could you point me to any sources? Sep 5, 2014 at 14:10
  • @ThomasGross Interesting - someone should do an acceptability study about this :) As for sources, sorry to disappoint you, but I would recommend asking on the phonet mailing list or perhaps someone from Phonetics in Cologne (they've worked on GToBi), if nobody else here comes up with any sources.
    – robert
    Sep 5, 2014 at 15:03

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.