Please note that the book is available in Europe from Elsevier Science Inc., Molenwerf 1, 1014 AG Amsterdam, The Netherlands, or in the USA/Canada from Elsevier Science Inc., PO Box 945, Madison Square Station, New York, NY 10160-0757, USA. Received April 1999) Reviewed by Sergio Zanini International School for Advanced Studies, Trieste, Italy David Birdsong ed.): SECOND LANGUAGE ACQUISITION AND THE CRITICAL PERIOD HYPOTHESIS. Lawrence Erlbaum, 1999. One of the more lively debates in SLA research concerns the explanatory adequacy of maturational constraints for the superiority of early over older learners in reaching native pro®ciency levels of a second language L2). Ever since Pen®eld and Roberts's 1959) proposal that the brain gradually loses plasticity, and Lenneberg's 1967) formulation of the Critical Period Hypothesis CPH), researchers have argued either for or against the existence of maturational constraints and a critical period CP) for language learning. During the late 1960s and 1970s the battle primarily concerned whether or not any advantages actually existed for children, the opposite sometimes being suggested on the basis of empirical data pointing to the superiority of older children and adults in early L2 development. However, after Krashen et al.'s 1979) review article, in which the pioneering division between `initial rate' and `ultimate attainment' studies was ®rst made, it became more or less accepted that even though older learners may have an initial advantage in the acquisition of morpho-syntactic aspects of an L2, ultimate levels of attainment seem to correlate negatively with increasing age of onset AO). The old layman observation that children are better language learners thus ®nally received scienti®c support. Consequently, the debate during the 1980s and 1990s has not focused on the existence of a young learner advantage, but rather on its causes. The anti-CP crowd has argued for non-biologically based explanations, suggesting instead socio-psychological factors including input and cognitive growth) as the major determinants for diering ultimate pro®ciency levels. On the other side of the fence, the CP proponents have relied on ultimate attainment data, such as those of Johnson and Newport 1989), appearing quite satis®ed with the fact that any serious threats to the CPH seem to be long in coming. However, despite the rich body of empirical studies, review articles/books, and edited volumes that have emerged during the past 30 years, this research ®eld has not made any signi®cant progress since the review of Krashen et al. 1979). We still seem to be at the point where we can only conclude that older starters sometimes outperform younger starters in initial REVIEWS 571