نشریه پژوهش های زبان شناسی
سال پنجم شماره 1 (پیاپی 8، بهار و تابستان 1392)

This paper aims to show that the phonetic differences in duration and spectral properties between stressed and unstressed syllables are small and f0 is dominantly responsible for signaling the position of the stress in Persian. The contrast arises in Persian through the existence of a rule which provides the final syllable of phonological words with a pitch accent, crucially skipping right-edge clitics. However, since right-edge clitics are not assigned a pitch accent, they are not included in the phonological word, instead forming a clitic group (CG) with the phonological word (PW) on their left. In order to investigate stress correlates in Persian, we composed a corpus of sentences featuring two minimal pairs contrasting a noun and a noun-plus-clitic combination: [tabeʃ] ‘light’ vs ['tab-eʃ] ‘toy+his’. The members of these minimal pairs were embedded in declarative and interrogative carrier sentences which varied across three focus conditions referred to as nuclear, post-focal and focal. A production experiment with 12 speakers yielded large f0 differences between the stressed and unstressed syllables, while it led to small differences in duration, spectral properties and intensity.

The present article deals with the description and analysis of structural deletion in Kalhori Kurdish. Optimality theory with a synchronic approach is adopted as the theoretical framework of this research article. The analysis of deletion process in this research paper consists of two sections. The first section deals with a mechanism by which deletion takes place and the second part is based on the analysis of the criteria of vowel selection for the application of deletion process. It is shown that Kurdish language should be included among languages which use both v1 and v2 deletion processes when applying hiatus, where a prefix and a lexical root coincide. It is also shown that deletion process, where two vowels meet, is not an optimal choice and that other processes such as glide formation can be more suitable here. It will also be argued that the deletion occurs only when a glide formation from v1 is not possible due to [Gh(GLIDEHOOD)] conditions.

The aim of this research is to study the vowel harmony in Azeri Turkish, spoken in Moghan region. It focuses on the role of opaque units in the process. Based on the generative phonology's linear approach, the vowel harmony is assumed to be the product of sequential assimilations of vowels in the string, whereas auto-segmental phonology considers harmonic features as being at a separate layer which is acting autonomously. For that, these features spread simultaneously on the whole vowels in the string, unless an opaque unit intervenes and blocks it. The analysis of the data collected from the region under investigation (namely from Germi city) showed that the vowel harmony can be described, according to the principles offered by Katamba (1989), as follows: a) vowel harmony could be observed in the features [±back] and [±round], b) [±back] harmony applies on the whole vowels in the system, while [±round] applies only on [+high] vowels, c) opaque units which block harmony feature's spreading are: 1) suffixes with a [+low] vowel in their underlying form and 2) the consonant /j/, occurring in the places other than root morpheme, which imposes [-back], [-round] and [+high] features on the preceding vowel, and this in turn influences the following vowels.

In speech recognition process, while listening to a word, a number of word representations are assumed to be activated simultaneously as candidates for an incoming word, and later the appropriate one is selected (McQeen et al. 1994). An important issue in speech processing is the kind of information that constrains lexical activation. Suprasegmental features are amongst factors which can play an important role in mental lexicon activation. The aim of the present study is to investigate the effect of lexical stress on mental lexicon activation in Persian. For this purpose, 10 Persian words with two possible stress patterns and two different meanings, and from different syntactic categories were selected. The target words were offered to the experiment participants in the form of three questionnaires. The results of the present study showed that: 1) stress pattern used while reading the target words is affected by the context in which the words are placed. 2) There is no relation between stress pattern used while reading the target words and the written word in front of them. In other words, when people want to write a word in front of the target words, they do not pay attention to the stress pattern they have used while reading the word. The reason of the second result could be that in Persian almost all verbs have just one stress pattern. Therefore, the participants consider the target words of this study as exceptions and do not pay attention to the stress pattern.

In this article, the noun-noun compounds collected from the Mo’in dictionary are investigated and 11 basic semantic functions (classes) are determined for them. The theoretical framework of this study is conceptual semantics (the semantic theory of the parallel architecture of grammar). In order to identify the semantic structure of these compounds, we should determine the semantic relationship between two nouns and find the semantic head of the compound. In this analysis, one noun can be an argument of the other noun or both of nouns can be an argument of another semantic function. This semantic function can be copied from the lexical entry of one noun or be one of basic semantic functions. In more complex situations, basic semantic functions can combine with each other. Therefore, a generative system is shaped, which can produce unlimited semantic relations between two nouns.

This paper deals with the phoneme substitution errors in children’s speech. The errors are felt by an adult listener as the replacement of one or more component sounds of a word by a different sound or sounds. The acoustic features of substitutive phonemes articulated by 19 girls and 18 boys of 2 to 7 years old within an acoustic space were analyzed and categorized using Praat spectrograph. The analysis of spectrographs and waveforms, the comparison of digital voice diagrams, and the calculation of percentage similarity of normalized Mel frequency Cepstral coefficients show that there is no reliable evidence for the concept of a substitution error in young children’s speech. In fact, some phonemic characteristics are neutralized in young children’s speech leading to audio errors in adult listeners’ cognition. The results are useful for providing an important metric of the child’s phonological development and for studying language learning theories as well as for curing children’s speech disorders.

Double object construction in the Persian Literary texts. This article reports an investigation of the double object construction (DOC) in a selective corpus of Persian literary works written from the Arab invasion (in ninth century AD) up to the present time. We also investigate dative alternation, the changes in structure and use of these constructions, and the possible sources of such changes. We propose that the use of the morpheme râ needs to be explained in reference to its roles as figure marker and within the framework of the Figure-Ground theory (Talmy, 1972). Another element of Persian DOC studied here is the morpheme mar which has been in use for more than six centuries. The proposal is that this morpheme has the same role as the figure marker in DOCs without râ, and in such constructions with râ plays a complementary role besides râ.