Topic and Focus Prosody in Russian - an Experiment
Gerhild Zybatow & Grit Mehlhorn (Leipzig)

Word order and intonation are relevant means of information structure in Russian.
Information structure is reflected in intonation phrases and is often correlated with syntactic constituents. We assume two different types of information structure: (i) focus-background structure and (ii) topic-comment structure (cf. Junghanns/Zybatow 1
997, Zybatow/Junghanns 1998). The underlying model contains the features [Top] (topic) and [Foc] (Focus), which can be contrastive [TOPc], [FOCc]. In speech production these features have articulatory and acoustic effects. They are realized at the surfac
e. Thus it is clear that intonation contours as suggested e.g. by Bryzgunova actually depend on placement of information structure features on syntactic constituents.

In order to find prosodic correlates of phenomena triggered by information structure we recorded spoken language and digitalized the recordings. Finally we carried out a thorough analysis of the resulting speech signals. We asked native speakers of Russi
an to pronounce everyday sentences to which they had had prior exposure. From the recorded sentences we extracted oscillograms, spectograms and intonation contours reflecting changes in fundamental frequency. The material included in the experiment consi
sted of Russian declarative sentences in their contexts, which made their information structure coherent and controllable.
Thus, we were able to make precise predictions about the assignment of information structure features. The sentences of the experiment varied with regard to word order and types of topic and focus (space prohibits including further details):

(1) prototypical topic, new information focus:

- (Kuda uexala Miroslava?)  Where did Miroslava go to?
- Ona uexala v JALtu.   She went to Jalta.
  [Top]      [Foc]

(2) contrastive focus:

- (Ljuda uzhe uexala v Jaltu?) Has Ljuda already gone to Jalta?
- MiroSLAva uexala v Jaltu.(Ljuda dazhe ne metchtaet o poezdke v
  [FOCc]      Krym.)
Miroslava has gone to Jalta. Lujda doesn|t even dream of going to the Crimea.

(3) contrasted topics:

MiroSLAva uedet v JALtu, a LJUda ostanetsja ZDES|.
[TOPc]          [Foc]      [TOPc]           [Foc]
Miroslava is going to Jalta, but Ljuda is going to stay here.

1. The main result is that Russian definitely has distinct prosodic patterns for topic and focus. The position and type of accent are characteristic of topic and focus.
Further results are:
2. Topics are realized with a rising-falling tonal sequence (low-high-low).
3. A rise of the fundamental frequency at the beginning of a sentence marks contrasted topics, so called I-topics. These topics are special in that the vertical amplitude is bigger compared with that of neutral topics.
4. A falling nuclear accent (HL*) corresponds to the natural focus. The exponent of natural focus in Russian is constituted by the last lexical accent, i.e. at the right periphery of a sentence.
5. Contrastive focus is distinguished by a rising-falling tonal pattern (LH*L), the vertical amplitude being bigger than the one found with neutral focus. The nuclear accent of contrastive focus accent is the most prominent accent of the sentence.

Thus, the analyses of the acoustic material serve as empirical verification of abstract syntactic and semantic assumptions encoded at PF. We took into account parameters like type and position of accent, level of fundamental frequency and time structure.
 
 
 
 

|BACK TO PROGRAM|