The thesis in this paper is that L2 speakers differ from L1 speakers in their ability to do memory storage and retrieval about linguistic structure. We would like to suggest it is possible to go farther than this thesis and develop a computational-level theory which explains why this mechanistic difference between L2 and L1 speakers exists. For this purpose, we believe a noisy channel model (Shannon, 1948; Levy, 2008; Levy, Bicknell, Slattery & Rayner, 2009; Gibson, Bergen & Piantadosi, 2013) could be a good start. Under the reasonable assumption that L2 speakers have a less precise probabilistic representation of the syntax of their L2 language than L1 speakers do, the noisy channel model straightforwardly predicts that L2 comprehenders will depend more on world knowledge and discourse factors when interpreting and recalling utterances (cf. Gibson, Sandberg, Fedorenko, Bergen & Kiran, 2015, for this assumption applied to language processing for persons with aphasia). Also, under the assumption that L2 speakers assume a higher error rate than L1 speakers do, the noisy channel model predicts that they will be more affected by alternative parses which are not directly compatible with the form of an utterance.

As pointed out by Carroll (Carroll), our team has investigated the influence of input on the spoken French competence of older Ontario bilinguals. Our research has examined the learning of invariant and variable aspects of French grammar. We focus here on the learning of variation, since it is an under-researched topic not covered by Carroll. Our research examines adolescent speakers of Ontario French from French-medium schools (e.g., Mougeon & Beniak, 1991), same-age immersion students (e.g., Mougeon, Nadasdi & Rehner, 2010) and advanced learners from a bilingual university (e.g., Mougeon & Rehner, 2015). Two key dimensions of input are teacher classroom speech and frequency of use of French in the community for the Franco-Ontarian students and amount of extra-curricular interactions with Francophones for the FSL students. Having collected corpora from these student groups, we compared the output of learners with primarily classroom-based input with that of learners with broader ranging (extra-) curricular input. The availability of teacher in-class recordings for these learner groups has been crucial in identifying additional factors influencing these students’ output.

We test the predictions of the Feature Reassembly Hypothesis (FRH) as applied to the L2 acquisition of French pronominal clitics by Anglophone learners, capitalizing on the fact that different semantic and morphosyntactic features are lexically encoded by French and English pronouns. A picture selection task and a self-paced reading task examine how the information encoded in the L2 forms affects off-line and on-line pronoun interpretation. Our findings suggest that the initial L1–L2 mapping was indeed influenced by the L1. Nevertheless, L2 learners successfully reassembled features into L2 bundles, as evidenced by target-like off-line performance. L2 reading time patterns, however, indicate that L1 representations may have a longer-lasting impact: learners’ reactions to the mismatching input followed a different pattern and were slightly delayed as compared to native speakers’. These results are in line with the FRH, which conceptualizes L1 influence as the transfer of atomic linguistic features and their combinations.

This study investigated the effects of L2 learning experience in relation to L1 background on hemispheric processing of Japanese pitch accent. Native Mandarin Chinese (tonal L1) and English (non-tonal L1) learners of Japanese were tested using dichotic listening. These listener groups were compared with those recruited in Wu, Tu & Wang (2012), including native Mandarin and English listeners without Japanese experience and native Japanese listeners. Results revealed an overall right-hemisphere preference across groups, suggesting acoustically oriented processing. Individual pitch accent patterns also revealed pattern-specific laterality differences, further reflecting acoustic-level processing. However, listener group differences indicated L1 effects, with the Chinese but not English listeners approximating the Japanese patterns. Furthermore, English learners but not naïve listeners exhibited a shift towards the native direction, revealing effects of L2 learning. These findings imply integrated effects of acoustic and linguistic aspects on Japanese pitch accent processing as a function of L1 and L2 experience.

In two cross-linguistic priming experiments with native German speakers of L2 English, we investigated the role of constituent order and level of embedding in cross-linguistic structural priming. In both experiments, significant priming effects emerged only if prime and target were similar with regard to constituent order and also situated on the same level of embedding. We discuss our results on the basis of two current theoretical accounts of cross-linguistic priming, and conclude that neither an account based on combinatorial nodes nor an account assuming that constituent order is directly responsible for the priming effect can fully explain our data pattern. We suggest an account that explains cross-linguistic priming through a hierarchical tree representation. This representation is computed during processing of the prime, and can influence the formulation of a target sentence only when the structural features specified in it are grammatically correct in the target sentence.

Carroll (Carroll) purports to contrast one view of language development in bilinguals, based on exposure, with another, based on processing. She claims that some researchers view language acquisition in bilinguals as founded solely on exposure and argues that language acquisition involves more than this, including analogy, associative learning, instance-based learning, rule-learning, and statistical learning.

Cunnings (2016) proposes that differences between native (L1) and second language (L2) sentence processing can best be explained in terms of susceptibility to effects of interference and an overreliance on discourse level cues during memory retrieval. Cunnings’ argument that difficulty in retrieval operations may provide a better explanation than a syntactic deficit account for explaining certain L1-L2 differences is convincing. However, the proposal for the ‘overuse’ of discourse is too broad and needs to be refined in terms of the specific contexts and conditions under which learners have difficulty. We also believe that difficulty with cue-based retrieval is still a characterization of the symptoms of differences between L1-L2 processing, and does not necessarily address the source of the variability.

The word frequency effect is stronger in second language (L2) processing than in first language (L1) processing. According to the lexical entrenchment hypothesis, this difference is not due to a qualitative difference in word processing between L1 and L2, but can be explained by differences in exposure to the target language: People with less exposure to a language show a steeper frequency curve for that language. Exposure differences can be measured with a vocabulary test. The present study tested whether the lexical entrenchment hypothesis provides an adequate explanation for differences in lexical decision times. To this end, we compared the performance of 56 Dutch–English bilinguals to that of 1011 English L1 speakers on 420 English six-letter words. In line with previous research, the differences in the word frequency effect between word processing in L1 and in L2 became vanishingly small once vocabulary size was entered as a predictor. Only in a diffusion model analysis did we find some evidence that the information build-up may be slower in L1 than in L2, independent of vocabulary size. We further report effects of cognates, age-of-acquisition, and neighborhood size that can also be explained in terms of differences in exposure.

In the current paper we report on a study of priming of variable Spanish 1sg subject expression in spontaneous Spanish–English bilingual speech (based on the New Mexico Spanish–English Bilingual corpus, Torres Cacoullos & Travis, in preparation). We show both within- and cross-language Coreferential Subject Priming; however, cross-language priming from English to Spanish is weaker and shorter lived than within-language Spanish-to-Spanish priming, a finding that appears not to be attributable to lexical boost. Instead, interactions with subject continuity and verb type show that the strength of priming depends on co-occurring contextual features and particular [pronoun + verb] constructions, from the more lexically specific to the more schematically general. Quantitative patterns in speech thus offer insights unavailable from experimental work into the scope and locus of priming effects, suggesting that priming in bilingual discourse can serve to gauge degrees of strength of within- and cross-language associations between usage-based constructions.

In this study we followed the characteristics and use of code-mixing by eight KODAs – hearing children of Deaf parents – from the age of 12 to 36 months. The children's interaction was video-recorded twice a year during three different play sessions: with their Deaf parent, with the Deaf parent and a hearing adult, and with the hearing adult alone. Additionally, data were collected on the children's overall language development in both sign language and spoken language. Our results showed that the children preferred to produce code-blends – simultaneous production of semantically congruent signs and words – in a way that was in accordance with the morphosyntactic structure of both languages being acquired. A Deaf parent as the interlocutor increased the number of and affected the type of code-blended utterances. These findings suggest that code-mixing in young bimodal bilingual KODA children can be highly systematic and synchronised in nature and can indicate pragmatic development.

In the present study, the Perceptual Assimilation Model (PAM) was tested on its applicability in child L2 lexical tone acquisition. The possible effect of L1 (Cantonese) lexical tones on L2 (Mandarin) lexical tone learning was explored. Accuracy rate and error patterns were examined with an AX discrimination task and a forced-choice identification task. Forty-nine native Cantonese-speaking students aged 8 years participated in the study. Results revealed that these children exhibited nearly perfect performance in the discrimination of Mandarin tones. However, significant tone differences were detected in the identification task. Tone 4 (T4) was identified with the lowest accuracy, and T1 with the highest. Error analysis revealed that Mandarin T2-T3 was the most confusing pair, followed by the T1-T4 pair. The inherent phonetic similarity between lexical tones in a language and the tone similarities across languages may also have contributed to perception difficulties, which could help to refine and supplement the PAM in the tonal/suprasegmental domain.

These studies address monolinguals' and bilinguals' processing of categories, in order to examine the relationship between concepts and linguistically encoded classes. We focus on languages that differ in their conceptual lexicalization and breadth of application, where one language has a single word (e.g., dedo in Spanish) that corresponds to two words in another language (e.g., English finger and toe). Categories differed across types of semantics-concept mappings, from ‘classical’ cases, involving members close in the conceptual space, to ‘homonyms’, involving conceptually distant items. Bilingual Catalan speakers, and English and Spanish monolinguals judged whether objects were ‘like’ an initial referent presented either with or without a label. Scores were highest in classical categories, lowest in homonyms; higher in narrow than wide categories; and better in labeled than unlabeled cases. Bilinguals outperformed monolinguals in judgments that conformed with their language, especially in wide categories. We discuss implications for the semantics-cognition interface and bilingualism.

Cunnings (Cunnings) offers an interpretation of L2-L1 sentence processing differences in terms of memory principles. We applaud such cross-domain approaches, which seem likely to significantly elucidate the neurocognition of language. Cunnings attributes sentence processing differences between (adult) high proficiency L2 and L1 speakers to an increased susceptibility to similarity-based retrieval interference, rather than to qualitative L2-L1 processing differences (cf. Clahsen & Felser, 2006). On his account, both L1 and L2 sentence processing depend upon a ‘bipartite’ working memory, which involves maintaining items active by focusing attention on long-term memory representations (Cowan, 2001).

Child language researchers have often assumed that progress in first language learning depends heavily on language exposure. For example, Hart and Risley (1995) compared children in middle class families with children in lower class families. Based on recordings made across several years in the home, they estimated that by the time the children from lower SES families entered first grade they had heard 30 million fewer words than the middle class children. Researchers and educators have argued that this ‘30 million word gap’ is a primary cause for academic failure of lower SES children in the primary grades in the United States. Researchers in second language acquisition (SLA) research have often postulated a similar linkage between exposure and attainment, both for early and simultaneous bilingual children and later second language learning. Carroll (Carroll) expresses justifiable skepticism regarding such claims regarding the effect of amount of exposure on language attainment. Despite some important differences in conceptualization of the nature of the input, I find her overall analysis compelling and important.

Idiom priming effects (faster processing compared to novel phrases) are generally robust in native speakers but not non-native speakers. This leads to the question of how idioms and other multiword units are represented and accessed in a first (L1) and second language (L2). We address this by investigating the processing of translated Chinese idioms to determine whether known L1 combinations show idiom priming effects in non-native speakers when encountered in the L2. In two eye-tracking experiments we compared reading times for idioms vs. control phrases (Experiment 1) and for figurative vs. literal uses of idioms (Experiment 2). Native speakers of Chinese showed recognition of the L1 form in the L2, but figurative meanings were read more slowly than literal meanings, suggesting that the non-compositional nature of idioms makes them problematic in a non-native language. We discuss the results as they relate to crosslinguistic priming at the multiword level.

In an impressive review of recent research on the L2 processing of syntactic and anaphoric dependencies and ambiguities, Cunnings (Cunnings) convincingly demonstrates that the evidence from studies conducted within the last decade is not compatible with the Shallow Structures Hypothesis (SSH) which posits that L2 readers engage in structurally less detailed parses than native speakers. He also concludes that invoking capacity limitations in working memory in L2 learners cannot account for the full range of findings. Instead, he puts forth a novel approach to non-native processing in the context of cue-based models of memory retrieval. Cunnings assumes (i) L2 learners are more susceptible to interference, and (ii) L2 learners weight retrieval cues differently from native speakers.

Eye fixation measures were used to examine English relative clause processing by adult ASL–English bilingual deaf readers. Participants processed subject relative clauses faster than object relative clauses, but expected animacy cues eliminated processing difficulty in object relative clauses. This brings into question previous claims that deaf readers’ sentence processing strategies are qualitatively different from those of hearing English native speakers. Measures of English comprehension predicted reading speed, but not differences in syntactic processing. However, a trend for ASL self-ratings to predict the ability to handle syntactic complexity approached significance. Results suggest a need to explore how objective ASL proficiency measures might provide insights into deaf readers’ ability to exploit syntactic cues in English.

Bilingual children are required to successfully develop phonological skills in two languages, yet little is known about the neural correlates associated with them. We obtained structural imaging data from 30 Hindi–English children aged between 8 and 10 years and used voxel based morphometry to explore neuroanatomical correlates of behavioural measures of phonological awareness. Our results showed that phonological skills in English are predicted by grey matter volume of bilateral putamen, but solely by right putamen in Hindi. Post-hoc analysis revealed that English nonword reading correlates with grey matter volume in bilateral putamen while in Hindi nonword reading it correlates only with right putamen. These differences in putamen-based mechanisms indicate that syllable level awareness sufficiently supports early literacy in the transparent, alphasyllabic Hindi orthography whereas that in English requires both phonemic and syllabic level awareness. Our findings point towards a key role for putamen in mediating phonological and reading skills in children.

This paper examines the ability of bilingual infants who were learning Dutch and another non-tone language to discriminate tonal contrasts. All infants from 5 to 18 months of age succeeded in discriminating a tonal contrast of Mandarin Chinese (Tone 1 versus Tone 4) and showed a U-shaped pattern when facing a less acoustically salient manipulated version (contracted) of the aforementioned contrast. Specifically, infants showed initial sensitivity to the contracted contrast during their early months, followed by a loss of sensitivity at the stage where tonal perceptual reorganization typically occurs, and a sensitivity rebound by the end of the first year after birth. Compared to a previous studying of ours testing monolingual Dutch infants (Liu & Kager, 2014), the discrimination patterns of bilingual infants revealed both similarities and differences. On one hand, as with monolinguals, non-tone-learning bilingual infants’ tonal perception presented plasticity influenced by contrast acoustic salience along the trajectory of perceptual reorganization; as well as a general U-shaped perceptual pattern when discriminating non-native tones. On the other hand, bilingual infants appeared to regain sensitivity to the contracted tonal contrast at an earlier age (11–12 months) in comparison with monolinguals infants (17–18 months). We provide several explanations, stemming from the simultaneous exposure to two languages, to account for the 6-month bilingual perceptual plasticity from linguistic and cognitive perspectives. The overall outcomes of the study offer insights into the infant perceptual reorganization and language development trajectory, expand on the differences between monolingual and bilingual language development, and broaden our understanding of the influence of bilingual exposure to the perception of non-native contrasts in infancy from linguistic and cognitive perspectives.

Carroll (Carroll) takes issue with the use of parent report to obtain quantity of language exposure measures in research on bilingual development. When discussing parent questionnaires, Carroll writes “Temporal units are crude measures of exposure and they tell us nothing about input”. While I agree that temporal units do not tell us much about the fine-grained details of the input within the temporal units, importantly, parent-report-based measures of input quantity have predicted variation in bilingual development of phonology, vocabulary and morphosyntax. These are robust and reliable findings across numerous studies, and yet, Carroll skates over them as if they did not matter, or dismisses them as trivial. Furthermore, Carroll seems to lead readers to believe that only coarse-grained, language-use temporal units have been obtained through this method; on the contrary, researchers have obtained fine-grained input quality details via parent report that also predict bilingual children's development. Finally, in some circumstances, parent report data is the only feasible method for obtaining language exposure information.