Although linguists often employ “native-language influence” as a shorthand for language transfer, most are well aware that crosslinguistic influence can come not only from the first language (L1) but also from the second language (L2) in cases where the target language is the third language (L3). Even so, there remains the question of just how similar the acquisition of L3 (or L4, etc.) may be to what is, strictly speaking, L2 acquisition. This question naturally intersects with the complexities of crosslinguistic influence, and the study of transfer beyond L2 has received increasing attention in the past 10 years or so. It was high time for a single volume that would bring together many strands of research on the problem, and this one edited by Cenoz, Hufeisen, and Jessner will no doubt become a key resource. The collection has an introduction by the editors and 10 chapters with the following authors: Cenoz; Hammarberg; De Angelis and Selinker; Ringbom; Dewaele; Ecke; Herwig; Kellerman; Fouser; and Gibson, Hufeisen, and Libben.

Most of the papers focus on lexis and on the acquisition of an Indo-European language. The study by Fouser is a remarkable exception on both counts, however, looking as it does at the acquisition of pragmatics of L3 Korean by two speakers of L2 Japanese and L1 English. Another exception is the article by Kellerman, who considers questions of gesture, emotion, and linguistic relativism; by his own admission, the article does not discuss L3 in much detail, but it is nonetheless an engaging look at problems of long-term interest for transfer researchers. The other papers address recurring themes in L3 research, including the role of language distance, the dynamic nature of multilingual competence, and the differing roles played by L1 and L2 or other languages known to the learner. It is possible to give a sense of the wide range of problems explored from considering a sample of the findings.

Two of the studies offer a contrast between the influence of two very different local languages on the target (which in both cases was English). In the study by Cenoz, the local languages were Spanish, which is an Indo-European language, and Basque, which is non-Indo-European. In the study by Ringbom, the languages were Finnish (non-Indo-European) and Swedish (Indo-European). Not surprisingly, the results are somewhat similar in the two studies, with the two Indo-European languages having a very different (and often greater) influence on learners' English. Language distance figures in other studies as well, two of which give cogent examples of morphological blends. De Angelis and Selinker, for instance, report an Italian interlanguage form è uccido “is killed” instead of the target Italian form è ucciso; the learner's L1 was English, but his L2, Spanish, no doubt explains the -ido suffix. Herwig offers similar examples involving multilingualism in Germanic languages. Contrary to the results noted so far, Gibson, Hufeisen, and Jessner did not find any straightforward effects of language distance in their study of prepositional verbs in L3 German, but the wide range of L1s involved in their study as well as other methodological problems will make it difficult to evaluate their conclusions.

If I had to pick the single most intriguing observation in this interesting volume, it would be one involving phonetics, even though the focus of the collection is on lexis. Hammarberg found that a native speaker of English used more L2 German pronunciations in her L3 Swedish at earlier rather than at more advanced stages of proficiency. As the learner became more fluent in Swedish, the phonetic influences from her L1 became stronger than the German. As Hammarberg argues, even an advanced proficiency in L2 (as was the case for this learner's German) seems to constitute a different kind of knowledge from the early established and highly automatic patterns of L1 phonetics.

Along with the factors already mentioned, transfer interacts with variables such as the formality of the speech situation, which is one variable foregrounded in Dewaele's study (with informal speech marked by more language mixing). Moreover, different methodologies give rise to different problems of interpretation, as seen, for example, in Ecke's elicitation of tip-of-the-tongue states. The multiplicity of factors that can affect L3 and subsequent languages makes it virtually certain that any comprehensive account of crosslinguistic influence remains a very long way off.