The languages of the world show wide variation in how they combine vowels and consonants. The basic unit of organization for vowels and consonants within words is the syllable, which is subdivided into onset and rime, containing the nucleus and an optional coda after the nucleus.
Languages vary in the combinatorial options for these units.
In theories of syllable structure, the concept of sonority (relative audibility) is central. The sonority of a sound places it on a scale with other sounds, the sonority hierarchy, from most to least sonorous. Syllables are assumed to be organized on the basis of this hierarchy, with the nucleus being the most sonorous sound in the sequence and sonority decreasing outwards towards the margins if there is an onset and a coda.
This approach is not without problems, as shown by recent discussions in the literature. Most studies on sonority, however, are based on case studies or a handful of languages.
We will contribute to this discussion in two ways, with a large-scale cross-linguistic comparison of patterns and a detailed case study of their acquisition. The resulting database of 500 languages will also be useful for other researchers.
The investigation of the acquisition of syllable structure will provide insights that advance research on disorders and developmental delay in children and in adults.
So far, we found out that there seem to be different sonority hierarchies active in different syllable constituents and different languages. The segment classes varying in sonority (sibilants, nasals and liquids) are in the center of the hierarchy and we assume that their variable location with regard to each other is determined by language-specific specifications of the contrastive features [±sibilant] and [±continuant] rather than by their phonetic properties. Considering word-initial consonant clusters we found that sibilant plus obstruent clusters behave differently both typologically and in acquisition from clusters of rising sonority, i.e., of obstruent plus sonorant. In connection with the investigation of consonant clusters we also looked at vowel epenthesis and vowel intrusion into such clusters and concluded that, besides syllables that can fill the nucleus with a consonant, there are also syllables without a nucleus. To account for our observations, we are working on a theory of the syllable, especially its left margin, that has a richer structure than known models. We hope to complete this picture by soon also looking into consonant clusters at the end of syllables and words.
In this project we intend to investigate the role of sonority, the sonority hierarchy and the sonority sequencing principle in the internal organisation of syllables. Preliminary typological findings by the applicant and a collaborator suggest that the current mainstream theory of syllable organisation is not well-founded empirically and in need of revision.
To investigate these theoretical issues and to be able to develop a more appropriate theory with a better empirical foundation we intend to create a database of the syllable phonotactics of 500 languages and to conduct research on children acquiring a language with complex syllable structure and a large consonant inventory as well as on subjects who lost competence in such a language to some degree, as in aphasia. Theories of phonotactic organisation have been developed and tested to date in individual case studies, or in typological studies with a very limited number of languages. Such studies always focus on very specific aspects of syllable structure or individual classes in the sonority hierarchy. While we appreciate in-depth case studies and narrow theoretical focus we have all reason to believe that these methodological approaches do not lead to appropriate scientific progress in the long run. We think it is necessary for theory development to keep the general picture in sight and test hypotheses on very different types of data (i.e., from acquisition, loss and cross-linguistic variation). The results as well as the data to be obtained in this project will be made publicly available to further scientific discussion of methodologies and development of theoretical models.
Since we intend to carry out research on acquisition and loss we expect that the results of this project will have practical repercussions for the wider population in the development of new diagnostic tests and therapies for language impairment based on the collected data and the theoretical insights gained in this project.