Null Stress in Polish and German, syntax based assignment?

Martin Prinzhorn (University of Vienna), Joanna Śmiecińska (School of English, Adam Mickiewicz University, Poznań)

The degree of integrity between the prosodic pattern of utterances and their syntactic structure has been the subject matter of much debate within the generative field; Chomsky' and Halle's (1968) Nuclear Stress Rule implicitly calls for syntax-independent stress assignment rules, Jackendoff (1972) proposes a set of complex linking rules between the prosodic and syntactic patterns of sentences. On the other hand, the discrepancy is viewed as only partial in Selkirk (1984) or Zubizaretta (1994) and most interestingly, as non existent by Cinque (1993), who in principle, calls for a totally isomorphic treatment of syntax and prosody claiming that the syntactic structure of a sentence alone is sufficient in specifying it's prosodic shape given his Null Theory of Stress.

Our presentation attempts to take a stance in this matter, based on the data from German and Polish.

First, we investigate German and Polish indicative sentences with (di-) transitive verbs where Cinque's system can be successfully applied and where, irrespective of the overt word order, the correct placement of the main stress is predicted, the context being that of the "neutral focus" in the sprit of Zubizarreta (1994), with the whole sentence qualifying as the focus:

1. What happened? (What's that nose?)

A boy has broken a window.

a) Ein Junge brach ein Fenster

a boy broke a window

b) Ein Junge hat ein Fenster gebrochen.

a boy has a window broken

c) Chłopiec zbił szybę .

boy broke window

d) Chłopiec szybę zbił.

Next, we discuss parallel ("neutral focus") intrasitive constructions in both languages as well as in English. Cinque's system implies a discrepancy between unaccusative and unergative intransitive verbs; whereas the placement of nuclear stress within the subject phrase in unaccusative constructions is expected (they qualify as "most deeply embedded elements" (Cinque (1993)) because of their deep structure object status) it will be problematic for unergatives, which are not mentioned by the author.

2. What's happening? (What's that noise?)

a) A baby is crying

b) Dziecko płacze.

c) Ein Kind weint.

Finally, an analysis of the placement of nuclear stress in wide focus interrogatives (direct yes-no questions) is carried out. Here, we also include Russian data, which offer an interesting contrast in comparison to English, German and Polish. An example is provided below:

3.

a) Do you like ice-cream ?

b) Magst du Eis ?

like you ice-cream

c) Lubisz lody ?

like2p.sing. ice-cream

d) Любишь мopoжeнoe?

like 2p.sing. ice-cream

In all the above examples the whole VP is in focus on an "out of the blue" wide scope reading. If these were to be changed into narrow scope readings it is for the verb and not the object in the Russian example to become the focus, contrary to English, German and Polish. The nuclear stress placement in (d) is thus very problematic for Cinque's Null Theory of Stress.

We propose a modification, but not a rejection of the syntax based nuclear stress assignment system, where the syntactic rationale behind a given stress pattern is present, it will not, however be reduced to the notion of "depth of embedding".

References

Chomsky Noam and Morris Halle. 1968. The sound pattern of English. Cambridge, MS: MIT Press

Cinque, Guglielmo. 1993. "A null theory of phrase and compound stress". Linguistic Inquiry 24(2):239-298.

Jackendoff, Ray S. 1972. Semantic interpretation in generative grammar. Cambridge, Mass: MIT Press.

Zubizarreta, María Luisa. 1994. Word order, prosody and focus. Ms., University of Southern California.