Search Results

Compared to more familiar varieties of Swedish, the dialects spoken in Finland have rather diverse syllable structures. The distribution of distinctive syllable weight is determined by grammatical factors, and by varying effects of final consonant weightlessness. In turn it constrains several gemination processes which create derived superheavy syllables, in an unexpected way which provides evidence for an anti-neutralization constraint. Stratal OT, which integrates OT with Lexical Phonology, sheds light on these complex quantity systems.

The chapter focuses on syllable structure and phonotactics, starting with Icelandic and discussing the motivation for assuming the syllable and subsyllabic constituents as structural entities. Among phenomena relevant in these considerations are the Icelandic length rule and the so‐called half length and overlength, both relevant to the way stress is realized. Faroese syllable structure is then discussed from the same point of view, and a template for the length rule in Faroese as it functions in full syllables is presented. Half length and overlength are also shown to occur in Faroese. There is a special discussion of the status of full syllables and restricted syllables in the phonological hierarchy. The remainder of the chapter is devoted to consonantal phonotactics, giving an overview of the permissible onsets and codas in each language. There is a special section on the analysis of gemination of glides and consonants.

Basic structural properties underlying the different categories of vowel prosthesis are enumerated and compared, revealing important characteristics shared by all categories. Of central importance is syllabic structure, and areas of further research into the syllable and its evolving architecture are identified. Finally, the significance of sociolinguistic forces as well as structural factors in shaping developments is reiterated.

Bert Vaux and Andrew Wolfe proposed a theory that syllabification is not exhaustive and that many consonants are left unsyllabified. They argued that unsyllabified consonants are appendices, or segments dominated directly by prosodic nodes higher than the syllable. This chapter argues for a theory of representation different from that of Vaux and Wolfe, one that still countenances unsyllabified consonants but arguably not dominated by any prosodic constituents. It first discusses the notion of prosodic licensing before presenting four alternative models of phonological representation. It then considers the implications of Eric Raimy’s (2000) theory for indicating precedence relationships and the role of the prosodic hierarchy, along with Cairns and Feinstein’s (1982) theory of syllable structure and the Sonority Sequencing Principle.

Chapter 2 discusses the Germanic and Vedic evidence for Sievers’ Law in considerable detail. This chapter is situated in Part One (Evidence for Sievers’ Law and the Possibility of Inheritance) the overall aim of which is evaluate the possibility that Sievers’ Law could have been inherited from Indo-European, even in principle. This chapter discusses pressing questions of relative chronology in Germanic. The possibility of a converse of Sievers’ Law in Gothic and Vedic is considered. Seebold’s Anschlußregel and Schindler’s restrictions on Sievers’ Law are discussed, in particular the moraic properties of obstruents (including laryngeals). The implications of Sievers’ Law for Indo-European syllable structure are considered. Edgerton and Lindeman’s arguments for word initial alternations and the importance of monosyllabicity as a criterion for alternation are set in the context of the formulaic language of the Rigveda and the potential skew which this could have produced in the evidence.

An alternative to the standard views of phonology is presented, to take into account the critical issues raised in the preceding chapters. Declarative phonology is polysystemic, non-segmental and abstract; it specifically excludes derivation and deletion as valid phonological mechanisms. Phonological structures are underspecified attribute-value matrices which attach to syllable structure at any level (syllable, onset, rhyme, nucleus, coda, or even higher, e.g. foot). Attribute-value matrices can be represented as underspecified tree diagrams. Phonetic implementation is exemplified and examples from Irish, Malay, German and Scots Gaelic are worked through.

This chapter starts by discussing aspiration in relation to the character of the fortis vs. lenis opposition, asking how the opposition can be represented and whether or how it is neutralized in certain environments. A phonological analysis is proposed relating preaspiration to other phenomena occurring in the same environments. It is shown that in some respects Icelandic preaspiration looks like being a separate segment, whereas the phonotactic distribution is restricted in that it only occurs before plosives. Preaspiration in Faroese shows similarities, but also differences from Icelandic. There is less reason to analyse Faroese preaspiration as a segment, rather it should be seen as a subsegmental property of the fortis plosives. The chapter ends with a discussion of the way preaspiration shows up in morphophonemics in both languages.

Chapter 1 describes the demography of the two languages and their historical relation, tracing them back to ‘Proto‐West‐Nordic’, the phonological characterization of which is based on evidence from Old Icelandic texts and partly on reconstruction. The West Nordic obstruents are often assumed to have had an opposition based on voicing, different from the modern languages. The West Nordic vowel system was rich in having nine vowel qualities, which in principle could take part in a length correlation and a relation of nasality. Old West Nordic had a number of diphthongs, as combinations of full vowel colours and semi‐vowels. The prosodic structure was of the type characterized as ‘moraic trochee’.

This chapter surveys some of the principal aspects of Romance prosodic structure. It covers the role of syllable structure in conditioning the distribution, quantity, and quality of stressed and unstressed vowels and the development of consonantal segments; phonotactic constraints; the role of the sonority hierarchy in segmental change; stress); length; and aspects of intonational patterns. Specific topics dealt with include: vowel quantity and syllable structure; consonant quantity; phonological processes and syllable structure (onset; s+C clusters; nucleus: vowels, diphthongs; coda); stress; rhythm; intonation (statements, narrow focus, yes-no question; wh-questions, imperatives).

This chapter lays out the possible consonant clusters and syllable structure, as well as various phonological processes in Oklahoma Cherokee. §3.1 discusses consonant clusters, and shows that Oklahoma Cherokee can have up to four consonants in a consonant cluster, subject to certain restrictions. §3.2 lays out the syllable structure of Oklahoma Cherokee. The syllable structure can be quite complex, and the onset especially can have a complex cluster. §3.3 will then discuss various segmental processes which remedy illicit clusters. Since vowel length in Oklahoma Cherokee is contrastive, both the mora and the syllable are relevant phonological units; §3.4 demonstrates that the levels of both mora and syllable are relevant in Cherokee, and some phonological phenomena refer to the mora, while others refer to the syllable. The syllable appears to be a more prominent phonological unit than the mora.

This chapter critically addresses some of the most important issues in Romance spontaneous and conditioned diphthongization. It covers ‘closing’ diphthongizations of vowels (notably /e/ and /o/) as a function of stress and syllable structure; the origins and development of new diphthongs arising through vocalization of syllable-final consonants, or through metathesis; the differential maintenance of the Latin diphthong AU and, in particular detail, the controversy surrounding the origins and nature of the ‘opening’ diphthongization of the Romance vowels /ɛ/ and /ɔ/, which are argued to lie in metaphony.

Language is a quintessential human capacity. In this chapter, we ask whether the human capacity for language relies on innate knowledge. As a case study, we examine the rules of phonology. This, seems like a curious choice. Most people believe they know how phonology works, and in their naïve theory, phonology requires no knowledge at all. Phonological patterns, people believe, are simply governed by the constraints of their ears, lips, and tongues. But a careful analysis of the anatomy of a syllable suggests this theory might be misinformed. These conclusions receive further support from an unexpected source—the phonology of sign languages. All these results suggest that our capacity for language could well be due to innate knowledge of abstract rules.

Chapter 1 concisely describes contrastive vowel length in Classical Latin, as seen for instance in mălus ‘bad’ vs mālus ‘apple tree’, addressing the different sources of evidence which allow us to establish its existence (internal phonological evidence as well as external evidence from orthography, metric poetry, metalinguistic judgements by Latin authors, comparative reconstruction within Indo-European, and evidence from borrowing from and into other languages). Particular attention is paid to the relationship of vowel length to consonant gemination (and, more broadly, syllable structure) through the different stages of the history of Latin. The chapter also focuses on the harbingers of loss of the vowel length contrast which can be spotted as early as prehistoric Latin, then becoming more and more salient at subsequent stages. It ends with a description of the structure and aims of the volume.

In Swiss German, which encodes a phonological contrast in consonant length, consonant duration signals the segment’s geminate status and, in medial position, indicates the word’s syllable structure. The present work investigates the interaction between these aspects of durational processing using the N400, an electrophysiological component which offers a fine-grained measure of the success of lexical access. A cross-modal semantic priming ERP study tested to what extent words with medial consonants whose duration had been phonetically lengthened or shortened (leading to an incorrect syllable structure) trigger lexical access. Behavioural and ERP results revealed a processing asymmetry: lengthening a singleton does not negatively impact lexical access, but shortening a geminate does. This asymmetry supports an underspecification account of the geminate/singleton contrast, and may indicate a bias towards initially parsing acoustic input according to a CV template.