The study also highlights the constructs of current linguistic theory, arguing for distinctive features and the notion 'onset' and against some of the claims of Optimality Theory and Usage-based accounts.
The importance of Henk Zeevat's new monograph cannot be overstated. [...] I recommend it to anyone who combines interests in language, logic, and computation [...]. David Beaver, University of Texas at Austin
EDITORS: Harry van der Hulst, Rob Goedemans & Ellen van Zanten TITLE: A Survey of Word Accentual Patterns in the Languages of the World PUBLISHER: Berlin: De Gruyter Mouton YEAR: 2010
Karen Steffen Chung, National Taiwan University, Taipei, Taiwan
SUMMARY This extremely ambitious volume attempts to survey the range of possible accentual system types -- “accent” and “stress” are mostly used interchangeably in this book -- in the world’s languages. The book is organized mainly geographically, by continent (e.g. Europe, Asia, Africa), region (Papua New Guinea and Irian Jaya, the Middle East), or language family (Austronesian); continent and family coincide with Australian.
The first chapter of Part I, “Word accent: Terms, typologies and theories,” by Harry van der Hulst, introduces the book’s approach, terminology, and the StressTyp database, from which the volume draws a good part but by far not all of its data. It clarifies the notions of and differences between accent, stress, pitch accent, pitch, and duration. Linguistic tone is specifically not addressed in detail in this survey, in order to maintain the book’s focus on accent/stress, but it is mentioned and described when it coexists with a stress system. It is in fact not always so easy to determine whether a feature should be considered “tone” or “accent,” and sometimes there is an interdependent and interactive relationship between the two, as in “hybrid” languages with a combined tone-accent system, of which there are many in Middle and South America. Often a language will however lean more to one side or the other. Sometimes the data cited by the authors is inconsistent with that in StressTyp, and this is generally pointed out in the text. With such an immense volume of data, there are bound to be errors, and evolving analyses; so one purpose of this project was to check through the StressTyp data and compare it with that cited in other sources.
Below are listed the topics and authors of the remaining chapters for Part I, with brief comments on each:
2. Australian Languages, by Rob Goedemans. Most Australian languages favor left edge as the location of main stress, though there is variation. This chapter also introduces the StressTyp database framework.
3. Austronesian Languages, by Ellen van Zanten, Ruben Stoel and Bert Remijsen. The Austronesian family number about 1,200 or 20% of the world’s languages, and they tend to favor right-edge stress.
4. Papuan Languages, by Ellen van Zanten and Philomena Dol. About 750 different languages are spoken in Papua New Guinea, some tonal, making it the most linguistically diverse area in the world, and this diversity also extends to stress types.
5. North American Languages, by Keren Rice. One classification posits 62 families and isolates, with expectedly diverse stress types.
6. Middle American Languages, by Harry van der Hulst, Keren Rice and Leo Wetzels. These are mainly divided into tonal and pitch accent systems.
7. South American Languages, by Leo Wetzels and Sergio Meira. The number of reliable descriptions for these languages is small, however, pitch accent and hybrid pitch accent-tone systems are common.
8. African Languages, by Laura Downing. Over 2,000 languages in four families are recorded for Africa in Ethnologue. The languages of sub-Saharan Africa are overwhelmingly tonal, often with stress as well. The well-described non-tonal languages are generally quite clearly stress languages.
9. European Languages, by Harry van der Hulst. This chapter covers languages in five families, extending to the Caucasus. Many types of accent systems are found, with some languages exhibiting tonal features, some with very weak accent.
10. Asian Languages, by René Schiering and Harry van der Hulst. This chapter covers a wide range of diverse families and languages, with many kinds of stress and tone systems.
11. Middle Eastern Languages, by Harry van der Hulst and Sam Hellmuth. This chapter is one of the shortest, since it only needed to survey languages of the Middle East, all Afro-Asiatic languages with stress systems, that were not already covered in “Asia” and “Africa.”
12. A typology of stress patterns, by Rob Goedemans. This chapter recaps the main features of accentual patterns of the world presented in the book. It could serve as a useful overview of the highlights and findings of the entire volume.
It is notable that the book starts with chapters on less-known linguistic areas and language families: native Australia, Austronesian, Papua New Guinea, and native North, Middle and South America. Likewise, some of the individual chapters start out with introductions of less familiar languages and language groups, e.g. Cushitic, Omotic and Nilo-Saharan for Africa, Chukoto-Kamchatkan, Ket, and Yukaghir for Europe, and Beja for the Middle East.
The different possibilities of stress placement in the world’s languages are presented starting on p. 32. Fixed primary accents can be counted from the left: initial, second, or third syllable; or from the right: ultimate, penultimate, antepenultimate. Variable accent may depend on syllable weight, or on lexical properties. Systems in which accent is assigned to one or the other syllable of a bisyllabic window are ‘bounded’. In accent systems with unbounded windows, primary accent can occur anywhere in the word.
Many of the authors start off their chapters by describing the survey’s limitations and inevitable omissions and errors -- but it is clear that all worked hard to produce as representative an overview as possible, carefully sifting through a mind-boggling amount of data and hand-picking the most illustrative bits, which they then organized, analyzed and commented on. Rather than just giving a linear account of what they found, it is clear that they digested the information thoroughly, synthesized it, and came up with many valuable insights on it.
In lieu of trying to provide an exhaustive summary, below are a few selected points in an effort to give readers some concise sense of the range of issues covered in this work.
1. Many languages, such as Swedish, have tonal contrasts only on stressed syllables; a stressed syllable is more “stretchable” in terms of expressing the features of duration, pitch, spectral tilt, vowel quality or clarity, and amplitude (pp. 5, 9).
2. There are differing views as to whether schwas – which for many languages must figure prominently in any discussion of stress – are lexical, or allophonic (p. 6).
3. Amplitude is not used distinctively in any language, and thus is only ‘allophonic’ (p. 7).
4. All languages allow monoconsonantal onsets (p. 7) – mentioned in a context that describes how stress is often linked to syllable structure and whether the syllable is “heavy” or “light”.
5. Stressed syllables may have extra phonotactic possibilities, i.e. greater complexity, may mark sites for morphological processes, or be anchors for intonational tones (p. 9).
6. The main difference between languages with stress and those with pitch accent is that pitch is the only main distinguishing feature in languages with a pitch accent system, in contrast to languages with stress, which is correlated with multiple features. Perhaps this should be obvious, but the difference between Japanese pitch accent and English-type stress might not be immediately apparent to language learners without clarification (p. 251).
7. In languages with contrastive length, duration is usually not the most important feature to mark stress (p. 325).
8. Some analyses claim that in some Tupí-Guaraní languages, nasality as an underlying feature always correlates with stress, so that there would always be at most one underlying nasal in a word, with other surface nasal vowels resulting from nasal spreading from the stressed vowel (p. 328).
9. Genuine non-stress accent systems seem to be very rare in South American languages; and this may be true of languages in general as well (p. 359).
10. In some languages, the penultimate or other non-final syllable may be preferred for tone, since a pitch drop in the final syllable may provide greater contrast; and spreading a tone over two syllables may make it more prominent than realizing it only on the final syllable (p. 414).
11. Compound noun stress in Germanic is touched on (p. 442) but, as in many other works on stress, is not treated in depth.
12. Tone has been found in dialects of German and Dutch (pp. 448-9).
13. Even fully tonal languages like Chinese are included in this survey. Beijing Mandarin has a trochaic stress pattern, due to the neutral tone, plus a system of alternating stress in polysyllabic words (p. 580).
14. Hebrew words are generally stressed on the final syllable … with the exception of children’s games; so: kla'fim is ‘cards’, 'klafim is a ‘card game’; mono'pol is ‘monopoly’, mo'nopol is the game ‘Monopoly’. This is also happening to personal names like Yael, and other words in colloquial speech, apparently influenced by the Yiddish accent (p. 624).
Discussion often draws on data from lesser-known languages, such as Alawa (Australia), Enggano (Indonesia), Siroi (Papua New Guinea), Polabian (West Slavic, extinct), Bawn (Sino-Tibetan), Murle (Sudan), and Siculo Arabic (extinct).
Part II presents brief profiles of the 511 languages from the StressTyp database as of 2009, with source references after each. The geographical areas are somewhat different from those in part I: Australian, Austronesian, Papua New Guinea and Irian Jaya, North American, Middle American, South American, Eurasian (all the languages of Europe plus those spoken in India that belong to the Proto-Indo-European language family), Asian, and Middle Eastern and African. This part is useful for quick reference on individual languages and language families.
EVALUATION It is hard to imagine any one, or even two or three people, having the expertise to take on entire chapters of a book like this with sweeping titles like “Asia” and “Europe.” Yet the book does an admirable job of what it set out to do. The book does not claim completeness, an impossibility given the current state of descriptions of accentual systems in many or most of the world’s languages, including even some widely studied ones. But it does lay a broad and solid foundation for descriptions of accentual systems in an impressive number of languages, making it a valuable contribution to our understanding of the possibilities of stress. The book can be considered “theory neutral,” making it easily accessible to linguists working in any framework -- a significant strength.
Being a thick book full of technical data, the volume doesn’t lend itself that well to cover-to-cover reading; and its main target purchasers are certainly first and foremost university libraries. However, scholars who specialize in prosody and stress in language might consider getting their own copy, slogging through it, and keeping it handy for reference. The rewards of doing so are many, as suggested by the preceding lists.
An organization based on a combination of geography, region and language family means that many unrelated languages are treated in the same chapter, while special concessions had to be made for languages spoken on more than one continent, e.g. Arabic could have been included in either or both Asia and Africa, but was in fact put in a special chapter, the Middle East. However, once readers familiarize themselves with the organizational scheme, it is relatively easy to find one’s way around. The subject index is not complete; for example, “segolate” isn’t listed, but this is a minor quibble. One could consider getting the electronic edition, or accessing it through a library -- this could solve the weight and bulk issue, and would have the added advantage of easy searchability. The book has a clean, attractive, easy-to-use format, with only a very few typos spotted -- no mean feat for a work of this scale.
_A Survey of Word Accentual Patterns in the Languages of the World_ might well be unrivaled thus far as *the* place to begin a study of stress in the world’s languages.
ABOUT THE REVIEWER
ABOUT THE REVIEWER:
Karen Steffen Chung is an associate professor of English and linguistics in
the foreign language department of National Taiwan University in Taipei,
and also teaches English over the radio and Internet. Her areas of
specialization include phonetics, teaching of pronunciation, and Chinese
morphology. She is the author of _Mandarin Compound Verbs_, which received
an NTU excellent research award in 2007, and is currently working on a book
on Taiwan English in collaboration with the Institute of Linguistics of the
Academia Sinica, Taipei.