Integration and Inference: Cross-situational word learning involves more than simple co-occurrences
Time: 12:10pm - 01:10pm
Place: Psychology 128
Statistical word learning involves forming and aggregating associations between words and objects that co-occur across contexts (e.g., Vouloumanos & Werker, 2009; Smith & Yu, 2008; Yu & Smith, 2007). However, the mechanisms that support such learning are currently under debate, including the extent to which learners carry forward multiple ambiguous associations (e.g., Trueswell et al., 2013). The current study presented adults with a set of statistical word learning tasks designed to measure what information from past contexts is available to further process and integrate with new information. Results reveal that learners use the co-occurrence of label-object pairings to make inferences both about objects and labels currently present and those presented on previous trials. Further, the strength of learners’ memory for past contexts moderated their inferences, suggesting a role for a rich information structure in cross-situational word learning.
In category: Child language acquisition
Spanish instructors' operationalization of task complexity and sequencing in non-experimental foreign language lessons
Time: 01:00pm - 02:00pm
Place: Woodburn Hall 002
One of the main goals of research on task-based and task-supported language teaching is to inform classroom practices and maximize opportunities for the acquisition of foreign languages. However, to date, much of the in-depth research has been limited to experimental or quasi-experimental designs, often lacking examinations of how ideas like task complexity translate to the foreign language (FL) classroom. Specifically, while there is rich literature detailing how tasks can be sequenced to maximize learners' cognitive engagement and subsequent language development (cf. Robinson, 2001, 2010, 2011; Robinson & Gilabert, 2007), to date, no study has investigated *instructors'* understanding and use of task complexity and sequencing in their own FL classes.
The current study extended task complexity and sequencing to non-experimental FL classrooms and was motivated by three research questions: (1) How do university-level Spanish FL instructors operationalize task complexity in their classes following training in task complexity and sequencing? (2) How do they sequence tasks? and (3) How do instructors interpret the outcomes of these tasks that they designed to be sequenced in terms of task complexity? Eight graduate student instructors in a task-supported language department at a large research university in the United States participated in the study; all were teaching beginning or intermediate-level FL Spanish. Importantly, the semester prior to the study, all instructors were enrolled in a semester-long FL teaching methods class, which included a course unit (three 75-minute sessions) on tasks, task complexity and task sequencing. In this TBLT unit, graduate student instructors discussed how to operationalize a task (e.g., Ellis, 2003, 2009; Samuda & Bygate, 2008), how to sequence tasks according to Robinson’s ideas on cognitive complexity (e.g., Robinson, 2001, 2010, 2011), heard a research-based guest lecture from an expert on task complexity and, as part of the training outcome, designed two tasks of increasing complexity to be implemented and sequenced in their Spanish FL classroom. The following semester, the researcher videotaped each of the instructors putting their tasks into practice in their Spanish FL class. An interview was conducted following each recorded lesson, targeting instructors' interpretation of task complexity, task sequencing, and task outcomes.
Results demonstrated that graduate student instructors’ task operationalizations aligned with TBLT research-supported ideas, as did their designs of tasks and task sequencing. There was large variation, however, in the specificity of how tasks were operationalized. Additionally, task outcomes were frequently interpreted in light of the linguistic outcome, which differed from the communicative focus of how tasks were operationalized. Implications for teacher training and pedagogy in this context, as well as future research projects examining how task variables translate to the FL classroom, are discussed in detail.
In category: Second language acquisition
Guampa: A toolkit for collaborative translation
Time: 03:00pm - 05:00pm
Place: Ballantine Hall 015
Alex Rudnick and Taylor Skidmore
In this talk, we will present Guampa, a package for collaborative translation meant to help both language activist/learner communities in generating resources for their heritage languages and MT researchers in building bitext corpora.
For most of the world's language pairs, large bitext corpora are not readily available, and would be difficult to construct. However, for some language pairs, such as that of Spanish-Guarani (the co-official languages of Paraguay), not only are there many speakers of both languages, there is a community of activists dedicated to the continued vitality of their heritage language. Guampa is a free software package for the online collaborative translation of documents. It includes tools for importing articles from Wikipedia and exporting bitext suitable for training MT systems.
In category: Computational linguistics
The development of lexical diversity in Intensive English Program Students in the US: Quantitative and qualitative measures
Time: 02:30pm - 04:00pm
Place: Ballantine Hall 205
Alan Juffs (University of Pittsburgh)
Second language (L2) learner corpora have become important because measures of frequency in learning are seen as a key to understanding L2 development (Ellis, 2007). One important measure of development is lexical diversity (e.g., Yu, 2010). Measures of lexical diversity can be compiled either independently of external lists or compared to frequency levels derived from standard reference corpora such as the British National Corpus (BNC) (Daller & Xue, 2007). This paper reports on a corpus study of the lexical development in the written work of 414 English as a Second Language (ESL) students in an intensive English program over a seven year period. The research question was whether their lexical diversity increased reliably over three semesters. Students contributed up to three semesters' worth of data to an online assignment submission system, resulting in a corpus size of 4,192,535 words. The texts were analyzed using a set of Python scripts, excluding non-dictionary words in the process. Sufficient data from Arabic-speaking (n=61), Korean-speaking (n=48) and Chinese-speaking (n=36) learners permitted tracking of lexical diversity over two to three semesters across the three L1 groups. A list-independent measure, D, shows that overall lexical diversity increases reliably over three semesters from low intermediate (average D score 51.2, SD=13.94) to intermediate (62.60, SD=12.68) to advanced (70.43, SD=20.83). Moreover, on a list-dependent measure of diversity, frequencies among words at the higher levels (> 3000 frequency band of the BNC) also increase, but this diversity measure in particular revealed qualitative differences across the three L1 groups. These differences are attributable to the culture-specific vocabulary used by the different L1 groups. We conclude that learners' vocabulary size and diversity increases significantly over time and that instruction in intensive programs should focus even more on academic words (Coxhead, 2000) to promote diversity at higher BNC frequency levels.
In category: Second language acquisition