This volume aims to rekindle interest in the question of whether insights from the study of second language (L2) acquisition can usefully be called on to provide explanations for pidgin and creole genesis, providing (mainly) creolist perspectives.1 Not surprisingly, the focus of several contributions is on transfer in pidginization, creolization, and L2 acquisition.

This article discusses how research on language transfer in the field of SLA can help to explain the origins of substrate influence in creoles and provide answers to more difficult questions concerning the distribution and verification of substrate features. First, it argues against the view that both SLA and transfer are not involved in the genesis of pidgin and creole languages. Then the view is presented that, as described in the SLA literature, transfer is not just a consequence of second language learning but also of second language use, and it serves as a communication strategy when the need arises. Such a strategy may be used by speakers of either a prepidgin or an already established pidgin when its functional use is being rapidly extended. Sociolinguistic perspectives on transfer in SLA, described next, throw some light onto the question of why substrate features remain in pidgins and creoles. Research on transfer in SLA also provides important insights into the specific factors that may have affected substrate influence in creoles. Evidence is presented that some transfer constraints discovered in SLA research—rather than other proposed factors such as so-called functional expendability—still provide the best explanation for the absence of particular substrate features in creoles. Finally, it is shown that tests proposed for verifying instances of L1 influence in interlanguage could be adapted for verifying instances of substrate influence in creoles.

This paper presents a study that attributes verb serialization in the interlanguage of Vietnamese-speaking ESL learners to language transfer and, furthermore, puts forward the view that such transfer bears a resemblance to substrate influence in creoles with serial verb constructions (SVCs). In a task that elicited English causatives through pictures representing the causation of events, a subset of the Vietnamese-speaking participants in this study produced a number of serial-type constructions that reflected lexicosemantic aspects of causative SVCs in Vietnamese. Speakers of Hindi-Urdu, a nonserializing language used for comparative purposes, did not produce any equivalents. Additionally, serial-type constructions with second verbs (V2s) representing a result (e.g., cook butter melt) predominated at lower levels of lexical proficiency, whereas serials with make and a result (e.g., make broken) were more evenly distributed across proficiency levels. One inference based on the results is that certain serials are eliminated early in the acquisition process through positive evidence obtained via English input, whereas others continue to appear beyond the elementary level because of misleadingly similar constructions in the input. A comparison of the proficiency-based transfer of “cook butter melt” serials in this study and the inferred transfer of SVCs in creolization suggests that, whereas transfer processes in the two contexts are congruent in certain ways (often resulting from the exigencies of communication, limited access to the TL, and linguistic convergence), the processes diverge because of differences in target norms and input conditions. The latter two factors provide one explanation for why SVC-related transfer effects were limited to a subgroup of Vietnamese-speaking participants in this study.

The advantages and disadvantages of wider or narrower definitions of pidginization and pidgin are reviewed to determine the differences between pidgins and naturalistically learned second languages (L2s). It is argued that a wider definition is preferred because it avoids problematic counterexamples and captures generalizations that allow us to view the difference between naturalistic L2 varieties and pidgins as one of degree, not of type. In first language (L1) acquisition, Bates and Goodman (1999) showed the link between the development of vocabulary and grammar and argued that this may be explained by, among other things, logical and perceptual bootstrapping. It is suggested that these types of bootstrapping are also relevant for explaining the pace of grammar development in pidgins and naturalistic L2 varieties. The tense-aspect system of a Spanish variety spoken by a Chinese immigrant in Spain is examined in detail. The data, taken from a 90-minute interview that yielded 602 tokens, reveal several clear traits of the informant's verbal system: (a) All nonfinite, imperfective verb forms (gerunds) correspond exclusively to Vendlerian activities; (b) all but three of the perfective nonfinite forms (past participles) correspond to telic verbs or predicates; and (c) 81% of the stative verbs appear in the third-person-singular present form. The sensitivity to aspectual distinctions in the informant's variety of Spanish is not addressed by logical and perceptual bootstrapping. Furthermore, although this sensitivity can be partially explained by language-specific considerations (i.e., transfer from Mandarin), such an explanation does not speak to precise form–function mappings found, which are best accounted for by appealing to the Primacy of Aspect and Distributional Bias hypotheses (Andersen, 1993; Andersen & Shirai, 1996).

In traditional classifications of languages by inflectional subsystems, both creole languages and the results of untutored SLA (interlanguages) are classified as isolating. We focus on remnants of verbal inflectional morphology in French-related creoles and ask: (a) Can the properties of verbal morphology be attributed to SLA, and (b) what does this imply for creole genesis? We show how acquisition of inflectional morphology occurs in learner varieties of French from a basic variety perspective. We discuss reflexes of French inflectional morphology that survived the creolization process, resulting in a long–short opposition of verb forms: Across creoles, different grammatical properties are marked by this opposition. Finally, we consider the implications for creole genesis theories and propose a three-generational scenario of creolization with multiple agents.