This book is part of the series The Phonology of the World's Languages, published by Oxford University Press. Even if the list printed as part of the book's front matter only comprises 18 volumes so far, all the North Germanic languages—Danish, Faroese, Icelandic, Norwegian, and Swedish—are now covered with the publication of this book.Footnote 1 The author is professor of Scandinavian languages at Stockholm University, and since the early 1990s, he has published a number of important works on Swedish and North Germanic phonology, both from a diachronic and synchronic point of view. His main interests are prosody and the interface between phonology and morphology, perspectives that are both well represented in the present book.
The book covers principally the variety spoken in Central Sweden, comprising the capital Stockholm. At different points, references are made to other regional varieties of Swedish, but a comprehensive and coherent account is only given of Central Swedish (CSw). The book is divided into 14 chapters. Chapter 1, Introduction, is an overview of how Swedish relates to other languages, the language situation in Sweden and in mainland Scandinavia, and a short description of official language policy. It also includes an overview of available language resources pertaining to Swedish. Chapter 2 introduces the phonemic and phonetic vowel inventory. Chapter 3 treats the consonant inventory and is organized in the same way as the previous chapter. Chapter 4, which contains an overview of the most important segmental processes in CSw, completes the set of chapters concerned with segmental phonology.
The following nine chapters treat prosody, beginning with the prosodic word in chapter 5. Riad distinguishes between a minimal and a maximal prosodic word, ωmin and ωmax, respectively, based on the recursive model of Itô and Mester (see, for example, Itô & Mester Reference Itô, Mester, Borowsky, Kawahara, Sugahara and Shinya2011). Both word levels are defined by culminativity: the minimal one by (one) stress and the maximal one by (one) tonal accent. The ωmax may be coextensive with the ωmin or contain more than one, as in compounds.
The analysis of stress is divided between three chapters, 6, 8 and 10, while tone is treated in chapters 9 and 11. The topic of chapter 6 is prosodic footing, which is claimed to be defective in Swedish in that only one stress foot is assigned per minimal prosodic word. Since a stressed syllable is obligatorily heavy, this foot takes the form of a moraic trochee. Chapter 7 looks at minimality and optimality of roots. What appears to be the canonical result of hypochoristics formation, that is, a branching rhyme (CVC) in the stressed syllable plus a monosyllabic suffix, is interpreted here as the canonical stem shape in Swedish. Chapter 8 discusses in greater detail what has already been mentioned at several points in previous chapters, namely, the requirement that stressed syllables be heavy. The dual manifestation of this requirement is referred to as complementary quantity: A stressed syllable has either a long vowel followed by a short consonant (which is extrametrical if final or forms the onset of the following syllable otherwise) or a short vowel followed by a long consonant or a consonant cluster. In the former case, the single consonant is underlyingly marked as moraic. In the latter case, the first consonant within the cluster is made moraic by the Weight by Position Principle, with some exceptions. This means that the quantity distinction underlyingly is encoded as a difference between moraic and nonmoraic consonants, and that vowel length is derived.
Chapter 9 introduces the contrastive tonal accents that characterize most varieties of Swedish and Norwegian. Riad is perhaps the main proponent today of the so-called privativity hypothesis, namely, that the difference between the two accents is that accent 2 has a tone as part of its melody, while accent 1 is lacking it. Otherwise, the tonal composition of the two melodies is identical. Another basic tenet of his analysis is that the distribution of the accents to a considerable degree is determined by morphology. (This is why it is hard to find “pure” minimal pairs distinguished by tonal accent. There is usually a difference with respect to morphological category that distinguishes a given pair in addition to the tonal difference.)
The following chapter presents Riad's analysis of stress placement in Swedish. It starts with the author taking issue with former analyses of Germanic stress, of which the majority has tried to derive the observed patterns primarily by means of phonologically based principles. Riad instead claims that stress placement in Swedish should be seen as a function of morphologically based constraints. He groups lexical items, stems, and affixes into four different prosodic types: unspecified, pretonic, tonic, and posttonic. Only with respect to the first type is stress placement a function of phonological principles. The three other types have stress placement encoded as part of their lexical specification, as implied by their names.
Chapter 11, which presents the structural analysis of the tonal accent system in CSw, draws a nice parallel with the preceding one. Just as affixes to a large extent determine stress placement, under Riad's analysis they also determine the tonal accent of the resulting word. In some plurisyllabic root morphemes, the accent 2 initial H, which is defined as the lexical tone, is part of the lexical specification of the root itself. However, in the majority of cases, this H is supplied by derivational and inflectional suffixes. This means that the H is part of the lexical specification of the suffix, even if it will associate with the stressed syllable of the stem.
Chapter 12, which to a considerable degree is based on Myrberg Reference Myrberg2010, presents the intonational system of CSw. Referring back to Bruce Reference Bruce1977, the basic hierarchical division is between focus accents and word accents. Focus accents head phonological phrases (PPs). Accordingly, each PP may contain only one focus accent. Word accents head maximal prosodic words. Each PP may contain one or more word accents. The intonational phrase is normally delimited by an initiality accent at its left edge and a final boundary tone, which may be low or high. The initiality accent has the same formal structure as a focus accent, although it does not express focus. The two final chapters are concerned with phono-tactics in monosyllabic words (chapter 13; intervocalic phonotactics being shortly treated in chapter 5), and the relationship between the phoneme inventory and orthography (chapter 14).
An important criterion of success that any book of this type must be evaluated against is whether it can be fruitfully read and exploited by phonologists of different theoretical persuasions. This criterion is certainly met in this book. The theoretical basis is generative phonology. All generalizations are stated as rules, without being overly formal. At different points, informal references are made to constraints, but no optimality-theoretic tableaux are offered. Throughout the book, data are presented in a clear, comprehensive, and systematic way.
This is not to say that all analyses are uncontroversial. Examples to the contrary are the analyses of stress and tonal accents. With respect to stress, the difference between Riad's analysis and earlier, more phono-logically based approaches, is not, in my view, as categorical as Riad appears to claim. In my opinion, the fact that stress placement is, to a certain extent, governed by derivational affixes is fairly uncontroversial and is taken into account by several of the sources cited at the beginning of chapter 10. Thus, the difference appears to be one of perspective. For Riad, phonological principles are invoked only when morphology is silent, while the opposite may be true for the earlier approaches he takes issue with. The common ground is clearly that both morphology and phonology are relevant and necessary elements of a comprehensive complete analysis of Germanic stress. That said, I must admit that I do not find the arguments in favor of morphology as the main conditioning factor offered on pp. 194–195 compelling. As far as I can see, the fact that speakers of Swedish find it relatively easy to adopt the original stress pattern of loan words with no apparent internal morphological structure is not an argument in itself in favor of morphology.
Riad's account of the tonal accent contrast is also controversial, but the alternative views are referred to in a way that I at least find satisfactory. Furthermore, even if I myself no longer believe that the privativity hypothesis represents the best way of analyzing the North Germanic tonal accents (see, for example, Kristoffersen Reference Kristoffersen2006, Reference Kristoffersen, Riad and Gussenhoven2007), this chapter is, in my view, a tour de force. First, it integrates the stress placement analysis from chapter 10 and tonal accent distribution into a coherent system. Second, and more importantly, it is, to my knowledge, the most detailed and comprehensive analysis of most, if not all, of the intricacies and irregularities of tonal accent distribution in Swedish (and Norwegian) to date. Thus, whatever one's view on what the best theoretical and analytical take on the distribution is, and whatever one's views on certain details of the analysis are, this chapter should become an obligatory point of reference for future work on North Germanic tonal accents because of its empirical scope as well as its completeness.
To conclude, I find that this book successfully fulfills its main purpose, which is to introduce readers with diverse interests and theoretical perspectives to the phonological system of the most widely used variety of Swedish, namely, CSw. In addition, many of the analyses in the book are important contributions in their own right to the study of Swedish and North Germanic phonology. The organization of the book may be somewhat unorthodox, in that two important features, stress and tonal accents, are both treated in several, nonconsecutive chapters. However, even if at first reading this may appear confusing, it highlights the parallels between the two systems, as discussed above.
