Vera Kempe

Processing lexical semantic and syntactic information in first and second language: fMRI evidence from German and Russian.

We introduce two experiments that explored syntactic and semantic processing of spoken sentences by native and non‐native speakers. In the first experiment, the neural substrates corresponding to detection of syntactic and semantic violations were determined in native speakers of two typologically different languages using functional magnetic resonance imaging (fMRI). The results show that the underlying neural response of participants to stimuli across different native languages is quite similar. In the second experiment, we investigated how non‐native speakers of a language process the same stimuli presented in the first experiment. First, the results show a more similar pattern of increased activation between native and non‐native speakers in response to semantic violations than to syntactic violations. Second, the non‐native speakers were observed to employ specific portions of the frontotemporal language network differently from those employed by native speakers. These regions included the inferior frontal gyrus (IFG), superior temporal gyrus (STG), and subcortical structures of the basal ganglia. Hum Brain Mapp 25:266–286, 2005.

THE ACQUISITION OF CASE MARKING BY ADULT LEARNERS OF RUSSIAN AND GERMAN

This study investigated the acquisition of the comprehension of overt morphological case marking by adult native speakers of English who were learning Russian or German as a second language (L2). The Russian case-marking system is more complex than the German system, but it almost always provides the listener with case inflections that are reliable cues to sentence interpretation. Two approaches to learning of inflectional morphology were contrasted: the rule-based approach, which predicts that learning is determined by paradigm complexity; and the associative approach, which predicts that learning is determined by the cue validity of individual inflections. A computerized picture-choice task probed the comprehension of L2 learners by varying the cues of case marking, noun configuration, and noun animacy. The results demonstrated that learners of Russian use case marking much earlier than learners of German and that learners of German rely more on animacy to supplement the weaker case-marking cue. In order to further explore the underlying mechanisms of learning, a connectionist model was developed that correctly simulated the obtained results. Together, these findings support the view that adult L2 learning is associative and driven by the validity of cues in the input.

Processing of Morphological and Semantic Cues in Russian and German.

This study examines the on-line processing of morphological cues to sentence interpretation in Russian and German with the goal of evaluating the relative impacts of cue availability and cue reliability. Both Russian and German use the cues of word order, animacy, case-marking, and subject-verb agreement to identify the agent of active transitive sentences. However, the availability of the case-marking cue is higher in Russian than in German. Using a picture-choice paradigm, we contrasted case-marking and animacy in Russian and German. The reaction times showed larger effects of casemarking in Russian than in German and effects of animacy in German, but not in Russian. These results suggest that the higher the availability of a cue, the larger the processing benefits associated with the presence of this cue and the smaller the impact of other converging information. A recurrent cascaded backpropagation network was designed to simulate these effects. The network succeeded in capturing the essential languag...

The Role of Learner and Input Variables in Learning Inflectional Morphology.

To examine effects of input and learner characteristics on morphology acquisition, 60 adult English speakers learned to inflect masculine and feminine Russian nouns in nominative, dative, and genitive cases. By varying training vocabulary size (i.e., type variability), holding constant the number of learning trials, we tested whether learners required a “critical mass” of vocabulary to generalize case marking patterns to new nouns. Cattell’s Culture-Fair IQ Test mediated the effect of type variability on success in generalizing case marking to new vocabulary: only participants with above-median CultureFair Test scores showed the predicted critical mass effect of better generalization with larger training vocabulary. These results demonstrate how individual differences in central executive functioning and attention allocation capacity can affect adult second language learning. Every natural language encompasses a highly complex system of categories at multiple levels of linguistic organization (Gomez & Gerkin, 2000; Lakoff, 1987). Language learners are faced with the task of organizing linguistic input in terms of phonological, semantic, pragmatic, lexical, morphological, and syntactic distinctions. This categorization problem is especially challenging to adult second language (L2) learners, who may be acquiring L2 categories based on limited input, and with potential interference from their knowledge of other languages (FrenckMestre, Foucart, & Caetano-Nunes, 2004; Gesi Blanchard, 1998; Hernandez, Li, & MacWhinney, 2005; MacWhinney, 1992). With respect to morphological categorization, languages vary enormously in the richness of their inflectional patterns. English, in particular, presents a very impoverished set of noun and verb inflections, relative to many other languages (e.g., Russian, Hungarian, Lithuanian, Finnish, Spanish). One feature of natural languages that is especially difficult for adult learners involves mastery of grammatical dependencies (Braine,

Crosslinguistic evidence for the diminutive advantage: gender agreement in Russian and Serbian children

Our previous research showed that Russian children commit fewer gender-agreement errors with diminutive nouns than with their simplex counterparts. Experiment 1 replicates this finding with Russian children (N = 24, mean 3;7, range 2;10-4;6). Gender agreement was recorded from adjective usage as children described animal pictures given just their names, varying in derivational status (diminutive/ simplex), novelty, and gender. Experiment 2 extends the gender-agreement elicitation methodology developed for Russian to Serbian, a language with similar morphosyntactic structure but considerably fewer diminutives in child-directed speech. Serbian children (N = 22, mean age 3;8, range 3;0-4;1), exhibited an advantage for diminutive nouns of almost the same magnitude as the Russian children. The fact that the diminutive advantage was found in a language with a low frequency of diminutives in the input suggests that morphophonological homogeneity of word clusters and membership in dense neighbourhoods are important factors that contribute to the reduction of inflectional errors during language development.

Structure emerges faster during cultural transmission in children than in adults

How does childrens limited processing capacity affect cultural transmission of complex information? We show that over the course of iterated reproduction of two-dimensional random dot patterns transmission accuracy increased to a similar extent in 5- to 8-year-old children and adults whereas algorithmic complexity decreased faster in children. Thus, children require more structure to render complex inputs learnable. In line with the Less-Is-More hypothesis, we interpret this as evidence that childrens processing limitations affecting working memory capacity and executive control constrain the ability to represent and generate complexity, which, in turn, facilitates emergence of structure. This underscores the importance of investigating the role of children in the transmission of complex cultural traits.

The role of sentence context in accessing partial knowledge of word meanings

Two experiments are reported that test the hypothesis that a reader can make use of the size of the semantic domain activated by a sentence context when inferring the meaning of a partially known word. We investigated words at three levels of knowledge: known, frontier, and unknown (e.g., Durso & Shore, 1991). Experiment I demonstrated that participants have knowledge about the meanings of words that they deny are part of the language (the unknown level), and that they make use of relative differences in the size of the semantic domains tapped by two sentences when asked to decide on correct usage of these unknown words. Experiment 2 demonstrated that participants have knowledge about the general semantic constraints operating on their unknown words, even when relative differences in size of semantic domains are controlled. Implications for the role of contextual constraints in vocabulary acquisition are discussed.

Individual differences in adult foreign language learning: The mediating effect of metalinguistic awareness

In this study, we sought to identify cognitive predictors of individual differences in adult foreign-language learning and to test whether metalinguistic awareness mediated the observed relationships. Using a miniature language-learning paradigm, adults (N = 77) learned Russian vocabulary and grammar (gender agreement and case marking) over six 1-h sessions, completing tasks that encouraged attention to phrases without explicitly teaching grammatical rules. The participants’ ability to describe the Russian gender and case-marking patterns mediated the effects of nonverbal intelligence and auditory sequence learning on grammar learning and generalization. Hence, even under implicit-learning conditions, individual differences stemmed from explicit metalinguistic awareness of the underlying grammar, which, in turn, was linked to nonverbal intelligence and auditory sequence learning. Prior knowledge of languages with grammatical gender (predominantly Spanish) predicted learning of gender agreement. Transfer of knowledge of gender from other languages to Russian was not mediated by awareness, which suggests that transfer operates through an implicit process akin to structural priming.

The crosslinguistic assessment of foreign language vocabulary learning

This study explores the use of task that can be applied in a uniform fashion across different languages to compare levels of vocabulary development in foreign language learning. Experiment 1 tested native speakers of Russian and German and demonstrated the basic comparability of subjects’ judgments for both words and nonwords. The results for Russian showed an influence of word length which can be understood in terms of the Orthographical Depth Hypothesis. Experiment 2 applied the same task to learners of Russian and German and found that learners of Russian had achieved a lower level of vocabulary control than learners of German at comparable language exposure levels. This disadvantage for Russian can be attributed to the novelty of the Cyrillic graphemic system which restricts the accessibility of written language input at early stages. There was a nonlinear increase over time in word sensitivity that can be attributed to the increasing contribution of lexical plausibility factors at later stages of learning. Also, the lexical decision task appeared to be sensitive to inhibitory effects of concurrently studied languages, as well as decay due to lack of regular exposure. Together, these results indicate that the lexical decision task can be a useful tool for the assessment and crosslinguistic comparison of lexical development in foreign language learning. Language teachers have long understood that some languages are more difficult than others. To properly assign students to languages at both universities and large language teaching institutes, it is important to understand the exact nature of differing challenges that languages present to the student (MacWhinney, in press). In order to better understand these issues, researchers interested in foreign language acquisition (FLA) need to perform crosslinguistic comparisons. However, these comparisons can only make sense if we can compare groups that are at equal levels of language exposure. For example, we might want to know whether second-year learners of German have learned more German than secondyear learners of Russian have learned Russian, given comparable levels of exposure? More specifically, we might want to know whether second-year learners of German have acquired a larger vocabulary or have more accurate pronunciation than second-year learners of Russian, given comparable levels of exposure. Vocabulary Assessment 2 If we want to conduct crosslinguistic experimental examinations of FLA, we need to find some way to equate level of attainment across learners of different languages. And we want to have ways of assessing proficiency that can be meaningfully interpreted in psycholinguistic terms. This paper explores a new methodology for developing such comparisons in the area of lexical learning and applies these techniques to a comparison of lexical learning in German and Russian. There are four reasons why current methodology needs to be adapted to provide a more solid basis for crosslinguistic studies. 1. The first problem lies with the common use of the duration of exposure as a surrogate variable for amount of exposure. Virtually all studies that make an attempt to assess different levels of learner ability share this weakness, whether they are conducted in the context of the Competition Model (Kilborn & Ito, 1989; McDonald, 1987), criticial period theory (Flege & Davidian, 1984; Johnson & Newport, 1989; Johnson & Newport, 1991), or Universal Grammar (White, 1990; White, 1991). The use of duration of exposure as a surrogate for amount of exposure is never defended on principled grounds. Virtually any standard theoretical perspective would assume that amount of exposure is more important than duration of exposure as a predictor of the degree of FL learning. However, there is good reason to believe that duration of exposure does not correlate tightly with the actual total amount of exposure. The amount of actual “time on task” is determined by the nature of the individual learner, the structure of the language program, the frequency of opportunities for interaction, and a host of other variables that are typically never measured or even examined. Thus, duration of exposure cannot reliably estimate amount of exposure. 2. Even if we had clear estimates of the amount of exposure to the FL, it would be dangerous to assume that similar amounts of exposure would lead to similar learning outcomes. Differences in learner backgrounds, the shape of target structures in the FL, possible L1 to FL transfer effects, and other factors can all lead Vocabulary Assessment 3 to important differences in the level of FL attainment that are not closely related to the amount of exposure to the FL. For example, learners of Russian must confront a new system of orthographic characters before they can even begin to read new words in the language. This will, in turn, delay the effective acquisition of new lexical items through textbooks and other printed sources. No such problem presents itself for learners of languages that use the Roman alphabet. 3. A third problem involved in providing a more solid basis for crosslinguistic comparisons involves access to crosslinguistically comparable proficiency tests. In principle, we would like to use proficiency test scores as covariates in crosslinguistic comparisons. However, solid proficiency tests only exist for high-profile languages like English, Spanish, German, and French. It is extremely difficult or perhaps impossible to find solid, recognized proficiency tests for even such important languages as Danish, Hungarian, or Vietnamese. Even when they are available, there is almost never a comparable form of the test for two different languages. For example, there are no comparable proficiency tests for German and Russian, or even for English and Russian. Paradis and Libben (1987) have developed an instrument for the study of relative levels of impairment in bilingual aphasics called the BAT (Bilingual Aphasia Test). However, this test can only be used to make comparative assessments of degree of impairment within individual subjects. It is not designed for use in group studies and cannot be used for crosslinguistic experimental work. In addition, it would be dangerous to attempt to make any direct application of a test developed for language-disordered bilinguals to groups of non-disordered nonbilinguals. 4. A fourth problem involves the mismatch between the measures provided by language proficiency measures and those needed for the prediction of outcomes in experimental studies. Proficiency tests typically include the assessment of a wide range of abilities including reading, writing, speaking and listening, as well as the Vocabulary Assessment 4 assessment of various language domains such as vocabulary knowledge, syntax, morphology and phonology. However, experimental studies of FLA usually focus on very specific aspects of language performance such as grammaticality judgments or reaction times in sentence comprehension tasks. If proficiency tests yielded a full set of subtest scores, we could use these individual components as predictors of outcomes in experimental tasks. However, often subtest scores are not available, or are not clearly associated with specific components of language learning. For crosslinguistic comparisons, we need measures which estimate the degree of the learners familiarity with specific FL skills without depending on measures of duration of exposure. These measures should be restricted to one specific language domain and should be easy to construct and easy to apply in a uniform fashion across languages. In this paper, we choose to focus our attention on the evaluation of lexical learning. We understand that a full crosslinguistic assessment of differences in target language difficulty will require the construction of multiple assessment insturments in each of the many subdomains of language performance(MacWhinney, in press). However, we must begin work toward this larger goal by delineating a specific, psycholinguistically manageable area for assessment. We have chosen the lexical domain for two reasons. Firstly, we believe that lexical learning is central to many aspects of early second language learning (MacWhinney, 1992). Lexical items provide an initial focus for learning in phonology, orthography, syntax, and semantics. Secondly, we believe that good experimental tasks are available for the crosslinguistic measurement of levels of lexical learning. One such task is the lexical decision task, which we have utilized in the current study. The Lexical Decision Task The task that we have chosen as a basis for our study is the standard lexical decision task that has been used so widely in psycholinguistics (Forster, 1976; Forster, 1985). In this Vocabulary Assessment 5 task, learners are presented with a large list of FL words and nonwords and have to decide for each item whether or not it is a word that they know. If they know the word they should say “yes”. If they did not know the word, they should say “no.” Meara and Buxton (Meara & Buxton, 1987) report an first version of this task, which served as the initial impetus to our work. One advantage of the lexical decision task over standard multiple-choice vocabulary tests is that the number of words presented in one session can be increased substantially, thus making the estimations much more reliable (Anderson & Freebody, 1983). The second advantage of a test based on the lexical decision task is that it minimizes the effects of strategies and contextual influences on test taking behavior. Anderson and Freebody (1983) carefully compared performance on a lexical decision vocabulary test and a multiple choice test, both containing the same test words, in 120 native 5th-graders. Although there was a strong correlation between the two tests, the authors showed that subjects were more likely to really know the meaning of words which were indicated as famili

Inconsistent handedness is linked to more successful foreign language vocabulary learning.

The study examined correlations between incidental learning of foreign words and interhemispheric connectivity, operationalized as consistency of hand preference, using pooled data of five experiments on adult foreign language learning (N=242). Inconsistent hand preference was found to be positively correlated with vocabulary learning even after effects of cognitive variables (verbal working memory capacity and nonverbal IQ), identified previously as predictive of successful foreign-language vocabulary learning, were partialled out. This observed relationship between handedness consistency and vocabulary learning persisted when left-handed and right-handed individuals were analyzed separately, and there was no overall difference in performance between left- and right-handers. The findings confirm an association between degree of handedness and verbal episodic memory.