The Interface of Intonation and Information Structure: The Representation of Degrees of Givenness in Persian Intonational Phonology

In studies on the realization of information structure in West Germanic languages (notably English, German and Dutch), it is commonly assumed that new information is marked by a pitch accent, while given information is deaccented (i.e. there is no pitch accent where one would otherwise be expected, see Cruttenden, 2006). However, a number of investigations of the prosodic marking of given and new information have gone beyond this dichotomy, taking into account different types of accentuation. One such study is Pierrehumbert and Hirschberg’s (1990) chapter on the interpretation of intonational contours in American English. They claim that both given and new information can be accented, and that it is the type of pitch accent which is used to differentiate between them (low: given; high: new). Two others within the British School (and on British English) are Halliday (1967a) and Brazil et al. (1980). Both, like Pierrehumbert and Hirschberg, allow for given information to be accented. Halliday (1967: 28) claims that contour type and accompanying local pitch range distinguish between given and new (mid-low to low: given; mid to low: new; and additionally high to low: contrastive new). Brazil et al. (1980:13) also claim that the nuclear contour distinguishes given from new but list different tones (falling- rising: given; falling: new). Furthermore, much of the work done on prosodic marking of information structure has concentrated on the binary distinction between given and new information, rather than different degrees of givenness. One exception is Pierrehumbert and Hirschberg (1990) who claim that a particular type of pitch accent can indicate whether information should be inferable, i.e. neither completely givennor completely new. However, most studies which refer to degrees of givenness are predominantly concerned with the morphosyntactic form of referring expressions. Only few combine morphosyntax and intonation, notably Allerton (1978) who nonetheless concentrates on accent placement rather than accent or contour type. The present research addressed the phonetic realization of degrees of givenness in Persian intonational grammar. The study particularly explored the question as to whether the F0 contour of a referring expression correlates with the degree of accessibility it is associated with in a discourse context. It was hypothesized that the F0 excursion size of a pitch accent correlates with the degree of givenness of the respective referring expression such that the more given a referring expression is, the less the F0 excursion size will be. The research methodology was the one used in laboratory phonology. A corpus of 10 small texts was designed to test the hypothesis. The speakers (12 male and 12 female) read the texts on a computer screen. They were instructed to read each text naturally, with no special emphasis on any part of the sentence. The target texts were recorded on DAT recorder using a high quality unidirectional head-mounted microphone (Shure SM58) in a sound proof booth. All the measurements were made on simultaneous visual displays of waveform, wideband spectrogram and f0 tracks. The results suggested that there is a negative correlation between the degree of accessibility of an expression and the pitch range or F0 scaling of its pitch accent; thus, the less accessible a referring expression is in a discourse context, the more compressed the pitch range of its pitch accent. The results further suggested that the pitch register of a referring expression significantly decreases as the expression becomes less accessible since degrees of givenness in the discourse context affects not only the F0 scaling the H tone but also the scaling the F0 valley. However, the tonal structure of a pitch accent is not affected by degrees of accessibility.