Five experiments were conducted on German topicalization sentences to explore their interpretation and the relation between the syntactic representation and the discourse representation of the sentence.

With respect to topicalization of objects, there are two main differences between English and German.

This so-called "split" topicalization is shown in (3a), with the NP part of the DP topicalized but the determiner stranded in its base position (see Fanselow and Cavar [2001] for a recent syntactic analysis of split topicalization and examples from a variety of languages).

Imagine a German topicalization of a phrase like einen Aufsatz ('a paper') in a sentence containing a quantificational adverb like frequently or several times.

The reason for these topicalizations are sentence-internal processing difficulties that would be associated with the adjunct in its basic position or any other sentence-internal position it might be used in.

I will call this change "topicalization." I will exclude cases of topicalization where the topicalized element is resumed in the sentence, that is, I will exclude cases of left-dislocation.

In German, topicalization refers to a position before the verb, left of the finite verb of the main clause.

Using the method of control paraphrases (see below), the project has compared English original texts with German translations, concentrating on sentences with felicitous topicalization. Section 2 will demonstrate the method and lay down the theoretical concepts involved: basic word order, information structure, language processing, and the principle of balanced information (BID).

The concept of felicitous topicalization is indirectly related to the concept of focus: as the head final verb phrase of German promotes end focus expectations, which the head initial verb phrase of English does not, less relevant or unfocused material precedes the focus in German and follows it in English.

Thus, for example, topicalization may secure end focus for the subject of the sentence, as in the example above, where the basic word order requires the subject before the verb phrase, and the adjunct verb phrase internally.

As the English version presents all the information, including the focus of the sentence, before the verb, topicalization of the adjunct does not improve focus identification.

(41.) Interestingly it turns out that topicalizations seem to behave on a par with HTLD in this respect:

With regard to some other LDs, Prince proposes that LD functions as a kind of strategy to circumvent an island violation in topicalization. To what extent these functions can be considered as additional discourse-pragmatic functions of LD is a question that deserves careful reconsideration.

My presumption is that a topicalization with a referential topicalized constituent can be syntactically realized as a covert LD-structure with an empty operator as originally proposed in Chomsky (1977).

On the limits of syntax, with reference to left-dislocation and topicalization. In The Limits of Syntax, Peter W.