Non-disordered acquisition of phonology tends to receive relatively little attention in general texts on language acquisition. The topic is given twelve pages in Guasti's (Reference Guasti2002) book-length language acquisition text, twenty in O'Grady's (Reference O'Grady2005) and a single chapter, out of nineteen, in Ritchie & Bhatia's (Reference Ritchie and Bhatia1999) child language acquisition handbook (Clark Reference Clark2009 is an exception, with much more extensive phonological coverage). Johnson & Reimers' contribution of an entire volume dedicated to the acquisition of phonology (and phonetics) is a welcome and important remedy for this gap.
The book begins with a substantial and data-rich chapter on recurring patterns in child phonological production. The patterns surveyed include reduplication, segment and syllable deletion, epenthesis (particularly of onsets and dummy syllables), featural modifications (voicing, stopping/deaffrication, place fronting, gliding) and consonant harmony. For each such pattern, a wide range of cross-linguistic examples is given, along with commentary on the universality (or the lack thereof) of the process among children. In addition, comparisons are made with similar processes in adult grammars, and their degree of attestation. This broad perspective is to be highly commended. Presentation of data first, then explanation and analysis, makes this section a valuable resource for classroom examples and problems.
The second chapter, on ‘Strategies’, is in part a continuation of the list of patterns, and in part a demonstration of how some patterns may combine into phonological conspiracies targeting certain outputs (e.g. simplified syllables). The discussion of consonant-cluster simplification is an example of the latter. It is followed by a revisitation of consonant harmony. Additional topics in this chapter are consonant–vowel interactions (discussed very briefly), velar fronting (to coronal) and segmental modification in different positions in the word.
The book continues with a chapter on linguistic models of language acquisition, which opens with an attempt to define markedness. The models which are presented are Jakobson's (Reference Jakobson and Keiler1941) framework, Stampe's Natural Phonology (1972) and then, jumping forward in time, Optimality Theory (Prince & Smolensky Reference Prince and Smolensky1993). The overarching goal of the chapter is to arbitrate between UG and non-UG approaches to language acquisition. In my view, the authors do little to acknowledge or discuss diversity of opinion and models within each camp. Jakobson's approach is dismissed as being too strongly predictive of universality, leaving OT as the sole representative of ‘UG’ models. It too is dismissed at the end of the chapter as being incompatible with the reality of early perceptual knowledge.
The development of perceptual knowledge is the topic of the next chapter. ‘The earliest stages’. Particular strengths of this chapter include comparisons between the abilities of non-human animals with human infants specifically, rather than human adult capabilities. The authors make thought-provoking observations about the validity of comparing trained adult primates and other animals, with varying levels of previous exposure to human language and to experimental protocols, with human children and adults. Finally, the authors relate linguistic development to general cognitive development in human children, though limiting themselves mainly to the topic of language lateralisation in the brain. A notable omission from this and the discussions of production is the phenomenon of covert contrast in child language, in which phonetic microvariation reveals phonemic differentiation in cases of apparent neutralisation, which goes undetected by adult listeners (for a recent overview of the sizeable literature on this phenomenon, see Munson et al. Reference Munson, Edwards, Schellinger, Beckman and Meyer2010).
These human/non-human comparisons set the stage for the discussion in Chapter 5 of ‘non-linguistic’ – i.e. ‘experience-based’ – perspectives on phonological acquisition. The authors include under this rubric scholars relying to at least some degree on exemplars, connectionist neural networks, variation in frequency and linguistic input, and statistical generalisation – some of whom might be dismayed to find their work described as ‘non-linguistic’. The result is to give the unfortunate impression of non-existent binaries regarding the roles of experience vs. innateness in language acquisition. For example, Kuhl's (Reference Kuhl2000) native language magnet model is the first discussed, which relies on the twin assumptions of a critical period and species-specific magnet effects, all the while disclaiming any role for UG. Later in the chapter, Johnson & Reimers hypothesise that human infants seem to ‘innately know what type of statistical information to look for or they make use of a learning ability other than the statistical one’ (pp. 113–114), but still without eschewing adversarial UG/non-UG, linguistic/non-linguistic terminology in favour of something more nuanced that can incorporate both experiential and innate – analytic and channel-based – learning (Nevins Reference Nevins2004, Moreton Reference Moreton2008).
The sixth chapter, ‘Towards production’, aims at synthesising the asynchronous development of perception and production into a single model. The authors review the problems faced by single grammars in accounting for this discrepancy, and introduce models with separate components for each domain, relying primarily on underspecification and structure-building. These models rely primarily on English data, in particular data from English-acquiring Amahl (Smith Reference Smith1973). Thus Johnson & Reimers' work highlights the need for available data on child production (magisterially reviewed in the first two chapters) to be incorporated into model-building as well. The chapter closes with a discussion of prosodic acquisition of syllable shapes and minimal words, which is fortunately able to be more cross-linguistic in nature (Dutch, French, Sesotho, Hebrew and Japanese data are present in addition to English).
The seventh chapter revisits patterns documented in the first and second, this time giving optimality-theoretic analyses of them. Consonant-harmony and syllable-simplification patterns receive the bulk of the attention. The book concludes with a chapter which summarises the preceding ones and offers the observations that both dual grammars (perception/production) and dual acquisition models (incorporating both experiential learning and some degree of innateness) are warranted.
Throughout the text minor copy-editing issues arise – for example, omitted function words on many pages in the first chapter. More important are organisational issues. Chapter 2's discussion of consonant harmony, for example could be more usefully be combined with Chapter 1's description of the phenomenon – consonant harmony is then revisited yet a third time in Chapter 7. Velar fronting is discussed in Chapter 2, separately from the general discussion of place-fronting in Chapter 1. Chapter 5 contains a clear and extremely useful overview of the physical development of the vocal tract and the segmental properties of infant babbling, oddly placed just after the discussion of non-UG models of language acquisition, separately from its natural home within a discussion of production development. In general, the interspersing of perception data, production data and comparisons of models is disjointed, while discussion of production processes is scattershot even within sections devoted to it. Finally, discussion of linguistic models is at times simplistic and marred by the problematic conflation until the very last pages of non-nativist, quantitative and ‘non-linguistic’ models with each other. While the authors do eventually introduce greater nuance in most cases, e.g. rehabilitating OT as capable of modelling both perception and production, this comes long after a theory's earlier dismissal, and too late for students reading excerpted chapters.
Students are indeed the primary intended audience of the book, which is being marketed as a textbook. It is suitable for students with at least one basic course in phonology/phonetics. While the basics of OT are reviewed in the text and a short glossary of phonological terms is provided, this may be insufficient for those with no previous exposure. The book would be appropriate as an assigned text for a language acquisition course, in conjunction with one of the general language acquisition textbooks mentioned above. Johnson & Reimers are able to provide considerably more detail on all aspects of phonological acquisition than Dresher's (Reference Dresher, Ritchie and Bhatia1999) or Fikkert's (Reference Fikkert and de Lacy2007) chapter-length and other (of necessity) data-thin overviews. For a seminar in phonological acquisition, it would function well as a stand-alone text or as a complement in breadth to an in-depth individual longitudinal study such as one of Smith's (Reference Smith1973, Reference Smith2010).
Overall, Johnson & Reimers provide an excellent and much-needed overview of child-language phenomena, especially in production. Particular strengths include the cross-linguistic data-richness and the comparisons with patterns in adult grammars, as well as the inclusion of data from animal and general human cognition. These features make it a valuable addition to courses in language acquisition and in phonology generally, as well as an outstanding reference and teaching resource for practising phonologists and acquisitionists.
