Since the publication of Clahsen and Felser’s (2006) keynote article on grammatical processing in language learners, the online study of sentence comprehension in adult second language (L2) learners has quickly grown into a vibrant and prolific subfield of SLA. As online methods begin to establish a foothold in SLA research, it is important that researchers in our field design sentence-comprehension experiments that adhere to the fundamental principles of research design typical of sentence processing studies published in related subfields of the language sciences. In this article, we discuss and review widely accepted principles of research design for sentence processing studies that are not always followed in L2 sentence processing research. Particular emphasis is placed on the design of experimental items and distractors, the choice and design of the poststimulus distractor task, procedures for presenting stimuli to participants, and methods for trimming and analyzing online data, among others.

There is presently a lively debate in second language (L2) acquisition research as to whether (adult) learners can acquire linguistic phenomena located at the interface between syntax and other modules, such as semantics, pragmatics, and lexical semantics, in contrast to phenomena that are purely syntactic in nature. For some researchers, the interface is precisely the place where fossilization occurs and the source of nonconvergence in L2 speakers. In this article we focus on the acquisition of the morphosyntax-semantics interface by examining the acquisition of mood in Spanish relative clauses by native speakers (NSs) of English. In particular, we focus on the contrast illustrated by Busco unas tijeras que corten “I am looking for scissors that cut-subj” versus Busco unas tijeras que cortan “I am looking for scissors that cut-ind.” When the indicative is used, there is a specific pair of scissors that the speaker is looking for. With the subjunctive, any pair of scissors will do, as long as it satisfies the condition expressed by the relative clause; the determiner phrase is nonspecific. In other words, we are dealing not with ungrammaticality, as both moods are possible in these contexts, but rather with differences in interpretation. General results showed that the learners could appropriately select the expected mood. We also saw that performance was not uniform across the various conditions tested. However, variability is not solely a product of L2 acquisition; we show it can be found in NSs as well.

As part of a research project on the investigation of second language (L2) ultimate attainment in 41 Spanish early and late near-native speakers of L2 Swedish, the present study reports on voice onset time (VOT) analyses of the production of Swedish word-initial voiceless stops, /p t k/. Voice onset time is analyzed in milliseconds as well as in percentages of word duration, thereby accounting for speaking rate effects. The results revealed an overall age effect on VOT production; however, this age effect became salient and statistically significant for all three stops only when speaking rate was taken into consideration. Similarly, when speaking rate was accounted for, only a small minority of the late learners exhibited actual nativelike L2 behavior, and most (but far from all) early learners performed within native-speaker range. The results are taken as an indication for relative VOT, as opposed to absolute VOT, constituting a reliable measure of nativelike L2 stop production, which has important implications for future research on age effects and maturational constraints in L2 acquisition.

This corpus study brings a second language (L2) research perspective, insights from generative grammar, and new empirical evidence to bear on a long-accepted claim in the World Englishes literature—namely, that inversion with wh-movement in colloquial Indian English is obligatory in embedded clauses and impossible in main clauses. It is argued that this register of Indian English is a L2 variety, functioning as part of a multilingual code repertoire, but that syntactic universals apply to first and second languages alike. Despite recent attempts at formalization, this distribution should be unattested, as such a grammar would fall outside the constraints of Universal Grammar and would contradict proposed discourse-pragmatic principles of natural language. A Perl program was created to search the Indian subcorpus of the International Corpus of English (Greenbaum, 1996) for relevant distributional patterns. Results reveal that wh-inversion in Indian English operates in the same way as in other varieties: It is robustly attested in main clauses and appears only occasionally in embedded clauses where syntactic and pragmatic conditions allow; it is obligatory only with interrogative complementizer deletion. Thus, contrary to the standard account but commensurate with recent corpus studies, users of English in India exhibit knowledge of universal constraints in this domain.

In this article, we challenge the notion that aptitude—operationalized as grammatical sensitivity as measured by the Words in Sentences section of the Modern Language Aptitude Test—is central to adult second language (L2) acquisition. We present the findings of a study on the acquisition of two properties of Japanese, head-final word order and case marking, by naïve learners of Japanese as a L2 who had no prior knowledge of this language or any other head-final, case-marking language. Participants underwent an input treatment in which they heard and saw basic subject-object-verb sentences in Japanese and were subsequently tested on these basic sentences and also on sentences to which they were not exposed during the treatment (i.e., polar questions and embedded clauses). Reading times on grammatical and ungrammatical sentences served as measures of underlying sensitivity to violations of grammatical word order for both the basic sentences and the novel sentences. Our results yielded three groups of learners: those who showed sensitivity to violations of basic word order, polar questions, and embedded clauses (parameter reset); those who showed sensitivity to violations of basic word order and polar questions only (partial parameter reset); and those who showed sensitivity to violations of basic word order only (no parameter reset). Because our measure of aptitude as grammatical sensitivity did not emerge as a factor distinguishing the three groups from one another, we argue that this component of aptitude is not a factor in whether learners are able to reset parameters. We also found that all participants, regardless of group, demonstrated sensitivity to case-marking violations, suggesting that aptitude as grammatical sensitivity plays no role in the acquisition of underlying features related to case or to the surface-level markings of case in Japanese.