1 Scope of the book

Introducing phonology is a beginning phonology textbook suitable for both undergraduate and graduate students.Footnote *

There are ten chapters, each divided into several sections, and each followed by a set of exercises (except Chapter 8) presented in order of increasing difficulty/complexity. Each chapter also includes the highly useful features of a preview with key terms and a brief summary in prose.

Odden introduces students to theoretical questions in phonology through data analysis. The book provides extensive and numerous data sets both for analyses provided within the chapter and for students to practise analysing on their own. With clear prose and ample examples and analogies, Odden explains and demonstrates various concepts of ‘classical’ (i.e. rule-based) generative phonology, with a particular emphasis on segmental phonology.

The textbook contains an astounding amount of data, some of which, if reviewed in its entirety, allows relatively broad exposure to a given phonological system, e.g. Korean and Kimatuumbi. It is also possible to develop an understanding of naturalness/markedness and typology of phonological phenomena by reviewing the types of segmental rules discussed in the volume (see also the comments on Chapter 8 below).

A large amount of space is devoted to phonetics and allophony (Chapters 1–3), which would probably be covered (in more detail) in a different, phonetics course (see further comments on this below). The remainder of the book is dedicated to the discovery of interacting and sometimes ordered segmental rules, with one chapter devoted to the autosegmental model of tone. As is familiar, current theory has largely abandoned the rule-based model in favour of constraint-based models such as Optimality Theory, so some might question the usefulness of a rule-based textbook. For novice students of phonology, however, the rule metaphor is intuitive. Furthermore, the concept of underlying vs. surface forms still plays an important role in OT. Knowledge of rule ordering will also prove to be valuable in a historical theoretical perspective as background to constraint ranking, as well as current controversies over opacity. Overall then, Odden's approach is motivated, as well as extremely well explained.

What is significantly lacking in the book is a discussion of syllable theory, a lack all the more acutely felt since Odden introduces the autosegmental model (for tone; ch. 10) and appears to assume that other issues, such as metrical phonology, the constraint-based approach and higher-level interactions (e.g. morphophonology) might be saved for a more advanced course. The absence of syllable-theory basics can lead to some confusion, since several segmental patterns presented (and many that could be/have been included) make reference to syllable structure, which is not introduced until §8.3, and there only briefly. Note that in Serbo-Croatian (ch. 7, ex. 1) one of the rules students are expected to find is stress placement on the last syllable of the word (ordered before separate C~V and V~ alternations); reference to syllables also arises in discussion of Paluan unstressed vowel reduction (ch. 4, p. 88) and Makonde (ch. 4, ex. 8). Notably, sonority is also not discussed. Tone, stress and vowel harmony patterns are treated along the lines of segmental rules, although, as mentioned above, the autosegmental model of tone is introduced in Chapter 10. A supplementary chapter on syllables is available online, along with some other materials (see below).Footnote ¹

Another general concern is that Odden's enlightening restatements of concepts are interspersed with lengthier exemplification – or are presented only in the chapter summaries – and thus are apt to be missed by student readers; e.g. ‘If rules apply in a particular order, you would expect to find languages with essentially the same rules A and B where A precedes B in one language and B precedes A in another: this is what we find in comparing Armenian and Lithuanian’ (p. 115). An occasional summary or key point in bold or italics would be beneficial. One nice feature of the text is the separation of advanced, tangential issues/problems that arise by putting them in ‘call-out’ boxes in the margins (akin to separation of text and footnotes) – though this is not consistent throughout all chapters.

Odden (2005) fills a notable gap in the area of textbooks on phonological theory. For comparison, the book covers fewer topics than Gussenhoven & Jacobs (Reference Gussenhoven and Haike1998) or Kenstowicz (Reference Kenstowicz1994), but emphasises data analysis, demonstrates phonological reasoning and motivates theory much more fully than its predecessors. At the same time, it is less encumbered by theory than Kenstowicz (Reference Kenstowicz1994), and has a much more accessible writing style.

2 Specific comments

Chapter 1, ‘What is phonology?’, is a brief clarification of the difference between phonetics and phonology, a necessary transition chapter for the beginning of a phonology course. Odden uses spectrograms and waveforms to demonstrate the continuousness of the phonetic signal compared with the assumption of discrete units (segments) for phonology. Explanation of the assumption – why it is an assumption and why the assumption is made – could be expanded.

Another crucial point made in Chapter 1 is the abstractness of phonology/ies, ignoring some phonetic details, or, in Odden's enlightening terms, ‘not all physical properties of speech sounds are cognitively important’ (p. 16; emphasis added).

One further clarification needed is that phonology, as a science, must then deduce (and not measure or prove) the units and structure of phonological systems. This point could follow from Odden's discussion of ‘theoretically possible word’ and ‘principled restriction of the language’ from §1.1 (p. 3).

In general, the writing is easy to read and is a good review of phonetics, but in this chapter, as well as in others, the key points need to be better highlighted. They could also be exemplified via a direct comparison of phonetic vs. phonological data: for example, phonetic ‘notable aperiodic noise around 3000–4000 Hz for 200 msec’ vs. phonological ‘this is a fricative’. The exercises at the end of the chapter make some of these comparisons.

Chapter 2, ‘Phonetic transcriptions’, is an overview of segmental transcription with discussion and an appendix of the transcriptional conventions of the International Phonetic Alphabet (IPA) and the American Phonetic Alphabet (APA). The chapter provides a review of articulatory phonetics and brief explanation of some (phonemic) distinctions (e.g. length) and other phonological details (e.g. stress, tone and syllabicity). Odden uses APA throughout the textbook, so the discussion and tables of the IPA/APA differences and the appendix are very useful for those with an IPA background.

Chapter 3, ‘Allophonic relations’, covers allophonic variation, elaborating on the reasoning and procedures likely provided in an introductory course. The concept of allophony is explained with reference to the classic problem of English stop allophones (flapping, aspiration and glottal stops) and morphophonological alternations (e.g. atom–atomic) – this usefully anticipates the morphophonological alternations emphasised in the remainder of the book.

Odden then walks the reader through the analysis of data sets from eight languages that exemplify various types of alternations, and provides one more set for the student to work through. This is likely too many for a student to read through all at once, and to grasp both the several clever and illuminating conceptualisations and the larger concepts, common phonological patterns and methods for analysing data. However, with reading guidance from the instructor and in-class discussion, the examples are highly useful. It would have been helpful for several sentences to have been highlighted or put in bold for the reader as key points and/or for discussion: e.g. ‘If there were no rule governing the distribution of consonants in this language, then the distribution would be assumed to be random …’. Eleven numbered exercises of varying difficulty follow the excellent chapter summary. To make use of the exercises here and elsewhere as homework, more complete instructions (and criteria for a write-up) would be helpful.

Chapters 1–3 are basically background chapters for the focus of the book, segmental rules. As such, those chapters might only be optional reading in a phonology class that follows a phonetics course, which makes three of the ten chapters of the book supplementary at best. For such cases, an inordinate amount of space seems to be devoted to phonetics, space that might have been better dedicated to syllable theory, for example.

Some phonetics courses I am familiar with do make use of Chapter 3, preparing students for phonology via allophony analyses, and Chapter 6, since it focuses on the articulatory basis of features. Though these chapters may be used in conjunction with other materials, Chapters 1–3 (and 6) would not be adequate reading and exercise material for an entire class introducing students to experimental phonetics.

Chapter 4, ‘Underlying representations’, embarks on the analysis of morphophonological alternations, which encompasses Chapters 4, 5, 7 (and 8). First, the concept of determining the underlying representation based on surface forms in a data set is demonstrated using Russian voicing neutralisation data. Next, the concept of an underlying form for each morpheme is explained in a dense but well-reasoned section, without data analysis. Next come four data-set analyses demonstrating how to find the UR, with special notes that (a) we might assume the UR is one of the surface forms (but that this may not always be the case), (b) the best UR is the one with the simplest explanation of (other) surface forms and (c) in formulating the rule, we should make it as general as we dare to, considering what the rule predicts for forms we don't have in the data set. These more difficult considerations are highlighted in call-out boxes in the margins.

Chapter 4 continues with four more data sets from Chamorro, Korean, Koasati and Kimatuumbi for the student to analyse. §4.5 demonstrates that the same word can show alternations within a sentence context, not just within a word context defined by affixation. §4.6 presents four more data sets from Samoan, Palauan, English and Tonkawa, which demonstrate that ‘it is wrong to think of deriving underlying forms by chopping off suffixes from some single column of data’ (p. 85).

The key term/concept of predictability might be better explained and integrated into the analyses in this chapter – the UR contains aspects of the morpheme that are not predictable (as outcomes of phonological rules). This basic fact is essential to determining underlying forms, and is a fundamental concept of phonology. Another useful addition would be consideration of the naturalness (in articulatory terms) or likeliness of the proposed rule in addition to the simplicity Odden mentions, since this chapter is dedicated to practising and considering basic segmental rules. Naturalness is discussed in Chapter 8 and rule likeliness in Chapter 6.

Chapter 5, ‘Interacting processes’, continues with the analysis of alternations, covering morphologically related data in sets that show more than one rule at work, sometimes interacting and sometimes crucially ordered. Starting with a data set from Votic, Odden demonstrates that rules should be simple – involving only one (featural) change (so, not /kə/→[ʃi] /Z#, but /ə/→[i] /Z# and another rule /k/→[ʃ] /Zi, where the vowel-change rule needs to happen first, creating the environment for the palatalisation rule).

Odden does not use the feeding/bleeding terminology in the textbook (although he does in the online supplement), terms which might have been introduced in call-out boxes. More importantly, the reasoning behind rule ordering (in general) is not discussed as explicitly as it could be – i.e. that one rule might affect the environment or target for another rule, and as such might create forms to which another rule applies as desired (feeding), remove forms from application by another rule (bleeding) or apply incorrectly (ordering is wrong). This apparent oversight is partly due to the lack of special formatting for key terms and important explanations and generalisations.

Odden does observe (p. 101) that ‘part of learning about phonology is learning what processes do exist in languages’ and goes on to discuss various types of rules in subsequent sections. There are three more data sets involving multiple rules (including post-nasal hardening, glide formation, labial elision, etc.). §5.2 involves data sets with crucial rule ordering, and §5.3 gives two complex data sets for student analysis/discussion, broken down into data subsets supporting a stepwise analysis.

This chapter is especially complex, in that students must learn about various types of rules (and what motivates them) as well as about rule interactions and how to deal with a larger data set. The data sets are arranged so as to lead the student through hypotheses and data to confirm or refute them step-by-step, in order to discover the correct analysis. Students would benefit from a review of the rules covered in Chapters 4 and 5, as well as some discussion of phonetic motivations behind them and a typology of rules (which is only suggested in Chapter 8) to bolster their confidence in hypothesising rules. Stepwise methods for analysis are also only suggested later, in Chapter 7.

Chapter 6 presents a set of features similar to those in SPE (Chomsky & Halle 1968), with explication of their definitions and functioning. This chapter seems oddly out of place, interrupting the chapters on analysis and failing to provide some background considerations for those analyses. Features are a good segue between phonetics and phonology, and are indeed often taught in an articulatory phonetics course. A chapter on features earlier in the text (after Chapter 1 or 2) could reinforce the focus on ‘cognitively important’ aspects of articulation as well as the larger goals of feature theory (and thereby cross-linguistic phonology), which are underemphasised in the book. The next logical step is allophonic variation (currently Chapter 3), introducing distinctive features vs. phonetic variation and redundant features. Features would again be needed for the formalisation of alternations, beginning with the current Chapter 4. It might then be re-emphasised in a later chapter that the best set of features will gain support from alternations and the natural classes they demonstrate. This interdependency between segmental rules and feature theory is likely why Odden orders the features chapter as he does. Given Odden's current ordering, discussion of feature theory is notably minimal.

Odden emphasises the phonetic differences between segments of different languages that nonetheless use the same transcription and the importance of phonology as a science in being able to predict which phonetic differences can be phonemic and what phonological rules are possible. The discussion of the features and their definitions (and how to go about changing the theory) is thorough and basically clear, but with the notable omission of inventories and how to determine the minimal set of distinctive features for a language – though minimal specification comes up in the discussion of natural classes in §6.3. Simplicity, ‘likeliness’ and formalisation of rules are clearly presented in §§6.4 and 6.5.

In my experience with this textbook, it is confusing to students that the feature matrices on different pages use different features and/or values – because the segments are from different languages: each matrix should be labelled clearly since they might be used for quick reference. Some typos are confusing, but are in the errata online. Other difficulties with this chapter are: [±stress] is given as a prosodic feature, as is [±high tone] – only the latter is discussed as problematic. [±syllabic] is also assumed, (English) /r/ is described as [—consonantal] and /f v/ are discussed as [+strident] (which would be wrong for English), in order to distinguish them from bilabial fricatives.

Chapter 7, ‘Doing an analysis’, is devoted to the further practice of segmental alternations and rule interactions with larger and more complex data sets. Odden emphasises hypothesis formation, testing and revision (and the gaining of experience with rules and problem solving) by walking the reader through the analysis of large data sets from five languages.

At about this time in a course, some students want a step-by-step method, and Odden suggests one in §7.1.2, ‘the first step: morphology’, an important point made earlier in Chapter 4, but worth reiterating here. The rest of the methodology, however, is not so clear-cut. The analysis of each data set has only three sections: the data, morphological analysis and phonological rules – or phonological alternations, the listing of which should be a step of its own in an analysis, preceding the postulation of any rules (hypothesis formation) and the testing and revising of postulated rules.

This is a difficult chapter for many students, but generally an important demonstration of how to do an analysis. The discussion of Japanese (so-called -u vs. -ru verbs), to take one example, is enlightening and interesting for those with a background in Japanese. However, it is very long and complex, involving thirteen often very specific (and rare) rules. This discussion may lead students to posit unlikely rules and complex rule sets, missing generalisations that are usually the goal in their practice data sets (e.g. the Serbo-Croatian exercise needs only three rules). Chapter 7 is more accurately ‘advanced data analysis’, rather than simply continued practice at the same level as Chapters 4 and 5. Still, a list of the typical steps taken in constructing a phonological analysis would be of help to most students.

Chapter 8, ‘Phonological typology and naturalness’, has no practice exercises, and its aim seems to be to provide a summary of the larger theoretical/cross-linguistic goals of segmental analysis, and to present a typology of rules. Odden discusses inventories and then types of assimilation, dissimilation, neutralisation and ‘prosodically based processes’. Introducing the concept of markedness, this chapter has the potential to help students gain an understanding of the typological basis of markedness (which guides most phonological analysis) and of markedness constraints, which are central to Optimality Theory.

The discussion of markedness needs strengthening, however, and what there is may be difficult for students to access. When markedness is introduced in the subsection on inventories, Odden writes: ‘This observation [that some sounds occur in more languages than other sounds] is embodied in the study of markedness, which is the idea that not all segments or sets of segments have equal status in phonological systems’ (p. 226; emphasis added). Far from being a definition of markedness, the sentence is confusing and vague (e.g. what kind of status?). The section continues with discussion of ‘common’ and ‘infrequent’ or ‘rare’ sounds, but with only occasional mention of ‘marked’ and ‘unmarked’; so the link between commonness and markedness may not be clear to students.

Moving away from markedness, the next sections explain types of rules. There are over twenty small, novel data sets exemplifying different types of rules in order to help the student understand rule types, but the novelty of the data sets can be distracting. A review of the rules in the data sets previously presented in Chapters 4, 5 and 7, data sets that the students are familiar with, would have been a good way to strengthen students' understanding of markedness as well as types of rule. Markedness could also be emphasised here by pointing out that common rules serve to ‘repair’ what would be marked structures. (This metaphor would also serve a constraint-based approach.)

The next section is devoted to prosodically based rules, including syllable-edge and syllable-weight phenomena as well as stress-based reductions and lengthening. (Here, Odden points out that many common prosodically based rules serve to avoid hiatus, but refrains from calling vowel sequences marked.) At this point, students are still unfamiliar with the basic syllable and stress facts, so this section is somewhat out of place. Finally, there is discussion of impossible and unlikely rules, which is valuable and, recalling discussion in Chapter 6, a much-needed review and summary.

Chapter 9, ‘Abstractness and psychological reality’, is a thorough and intellectually challenging discussion of phonological theory via abstractness in underlying representations and ‘what counts as evidence for a phonological analysis’. Odden explains and demonstrates the theoretical concepts behind SPE and post-SPE types of segmental analyses. Two cases of ‘well-motivated abstractness’ (from Kimatuumbi and Sanskrit) show that there is sometimes only indirect evidence for a UR that is nonetheless well motivated. A useful but advanced discussion of an unmotivatedly abstract UR of English ([ɔj] as /œ/) follows, with a discussion of what morphological relatedness is and of the importance of productivity (this should be a key term) of morphological processes in determining morphological relatedness (as the backdrop to phonological rules). Additional sections discuss abstract analyses of Maltese (Brame Reference Brame and Brame1972) and Yawelmani (Kisseberth Reference Kisseberth1969), Kimatuumbi and a Bedouin Arabic language game (where productivity could be reiterated).

These are deep and controversial issues in phonological theory, and as such, this is a conceptually advanced chapter. Odden provides ample data from a number of languages to support and demonstrate the concepts, and his prose is clear and incisive. Conceptual questions for discussion would be very helpful as chapter review exercises: e.g. what is productivity?; what is psychological reality and how does it relate to phonological analysis?; how are alternations good evidence for a posited UR?

In a section of Kiparsky's (Reference Kiparsky, Bach and Harms1968) use of historical change as a theory-external test of theories, several advanced/key terms are highlighted in coloured text for the first time in the book. This is the type of highlighting that is needed throughout – for key terms and key conceptualisations – to make this book textbook-like and accessible for the beginning phonologist.

The final chapter, ‘Nonlinear relations’, is an introduction to the autosegmental model via tone. The introduction demonstrates the motivation for the reconceptualisation of phonological theory (from linear to autosegmental model of tone) by pointing out the failed predictions and arbitrariness of the linear model vs. the naturalness and principled explanations provided by the autosegmental model of contour tones. Tonal stability, tone-melody inventories, tonal polarity changes and floating tones are presented as further evidence for the autosegmental analysis, with data from African tone systems and Mixtec. Tonal morphemes introduce mapping directionality, followed by toneless vowels and default tone, and tonal mobility. Odden then extends the autosegmental conceptualisation to features with data on [+high] in Vata and [—continuant] in Fula, followed by lengthier discussion of [constricted pharynx] in Aramaic. Finally, Odden briefly presents basic feature-geometry concepts before the chapter summary and (only) two exercises on tone analysis (but more are available online). There could be many more short exercises to get students used to testing tone analyses by manually drawing out the tone-to-TBU mappings as a way to provide evidence for their analysis.

A review and conclusion chapter with conceptual questions as exercises could also be of use.

3 Other features of the book

A glossary of phonological terminology precedes the references and language and general indexes.

A good deal of supplementary material is available online, including: errata; data from the exercises of Chapters 2–5, 6–7, 9–10 (for cutting and pasting); a chapter with extended analyses of Kimatuumbi and Syrian Arabic phonologies (multiple rules and interactions); APA symbols guide; definitions sheet for features; ten additional tone problem sets; a few additional data sets (for Chapter 7); and assorted other extensions of content for Chapters 1–6, as well as a chapter on ‘Prosody’ (covering length, skeletal slots, syllables, feet and stress).

The section on syllables in the online chapter on prosody underplays the motivation for the syllable in phonological theory. ‘The main argument for the syllable is its phonological utility, specifically the fact that generalizations can be stated more efficiently by reference to the syllable’, Odden writes, giving the impression that syllables are merely expedient for phonologists, when in fact, simplicity is a main reasoning factor in Odden's discussion of segmental rules, and underlies Occam's Razor, which guides most scientific reasoning. The online chapter continues with several arguments for the syllable, mainly from English, presented with minimal data (far from his analysis of large data sets for segmental alternations), and obvious scepticism, which can strongly influence students at this level. There is no data analysis, minimal discussion of languages other than English, no discussion of sonority and only a brief introduction to (traditional, premoraic) subsyllabic constituency. Odden then moves on to stress and the foot, with a section on quantity-sensitive stress, followed by non-stress uses of the foot (e.g. minimality). Thus the online chapter is far from a complete presentation of prosodic structure and its analysis.