The processing of number and gender features in Heritage Spanish: Insights from eye-tracking

• Zuzanna Fuchs, Ph.D., University of Southern California

Work on agreement in Heritage languages has observed that Heritage Speakers’ production and comprehension of agreement – particularly agreement in grammatical gender – differs from that of baseline speakers. For Heritage Spanish in particular, it has been suggested that this may reflect a different representation of number and gender in the heritage grammar. The current project uses eye-tracking in the Visual World Paradigm to investigate Heritage Speakers’ real-time processing of number and gender features. Building on prior work that shows Heritage Speakers of Spanish show target-like facilitative processing of gender features on definite articles el and la, we now turn to consider how they process articles that contain cues not just to gender but also to plurality (los, las). The goal of this talk is to demonstrate how evidence from real-time processing can complement research using offline tasks to enrich our understanding of certain mechanisms in Heritage Languages.

Measuring heritage speakers’ grammatical knowledge using acceptability judgment tasks

• Bradley Hoot, Ph.D., DePaul University

This presentation reports the results of an experiment examining syntactic island effects with heritage speakers of Spanish via an acceptability judgment task (AJT), with special emphasis on the methodological considerations involved in employing AJTs for investigating the grammatical knowledge of heritage speakers. Syntactic islands are structures in which long-distance dependencies are not possible (e.g., * What did you hear the rumor that Mateo copied ___?). They have long been of interest to linguists but have mostly been studied with monolinguals. In this experiment, 183 heritage speakers of Spanish in the US rated two island structures on a seven-point numerical scale, and the results suggest that their knowledge of these islands does not differ from that of the baseline. I discuss the methodological and analytical decisions we made to reach that conclusion, including operationalizing the syntactic effect, designing the items, sampling, data processing, statistical analysis, and interpreting the results.

Toward understanding the characteristics of heritage accent

• Ji Young Kim, Ph.D., University of California, Los Angeles

Over the past few decades, research on heritage language phonetics and phonology has grown tremendously. Although heritage language speech sounds appear to be quite similar to those of homeland varieties, heritage speakers are often considered to have an accent that is distinct from the accent of homeland speakers. In this presentation, I will discuss experimental research methods in heritage language phonetics and phonology. Particularly, I will discuss bottom-up and top-down approaches to the understanding of heritage accent. Phonetic features contributing to heritage accent can be narrowed down by first identifying areas of divergence from homeland varieties and then examining whether divergence in those areas leads to perceived heritage accent (i.e., bottom-up approach). Characteristics of heritage accent can also be identified by conducting holistic judgment on resynthesized speech that is designed to tease apart the effects of individual phonetic features on perceived heritage accent (i.e., top-down approach).

Different types of vocabulary knowledge: Comparing meaning recall and meaning recognition in heritage language bilinguals

• Eve Zyzik, Ph.D., University of California, Santa Cruz

This presentation examines two different formats for measuring vocabulary knowledge: meaning recognition and meaning recall. In research with second language (L2) learners, it has been shown that meaning recall tests are generally more difficult, and current debates center on whether these formats tap different psychometric constructs. I extend this line of research to heritage speakers of Spanish in order to determine the degree of overlap between the two test forms. A set of fifty lexical items were chosen for testing, including high frequency (n=13), mid-frequency (n=32), and low-frequency words (n=5). Participants took two vocabulary tests with a time interval of three weeks between testing. They first completed a meaning recall test, with the words embedded in non-defining contexts. The second test was meaning recognition, with five multiple-choice options, including an “I don’t know” option to prevent random guessing. In addition to the quantitative analysis, I examine words in which there is a mismatch between test formats (e.g., words known in recognition, but not recall). It is likely that these mismatches are due to educated guessing behavior that may be unavoidable with multiple-choice options on tests. I conclude the talk with general considerations for testing vocabulary among heritage speakers for proficiency or placement purposes.