This proposal requests funding to launch a research program to understand how and why speech patterns vary across neighborhoods, as well as the implications of speech for the schooling outcomes of disadvantaged children. Speech is socially constructed, and so neighborhood environments may have some effect on the use of dialects such as African-American Vernacular English (AAVE). Previous research suggests use of AAVE is negatively associated with academic achievement, which could occur if AAVE makes it more difficult to read and write standard American English (SAE), or because use of AAVE may engender discrimination from teachers. We have assembled a team of leading economists and socio-linguists to address these questions by exploiting unique new speech data collected using comparable speech measures as part of the HUD-funded Moving to Opportunity (MTO) randomized mobility experiment, and as part of a major population study, the National Longitudinal Survey of Youth, 1997 cohort (NLSY97). We request funding for the following main aims: 1. Transcribe speech samples from MTO and NLSY97, categorize speech through listener perceptions of the speaker's race, as well as linguistic analysis, which will include measurement of phonological, morphosyntactic, phonetic, and acoustic features; 2. Document the amount of across-neighborhood variation in speech that exists in the national NLSY97 sample, and determine how much of this variation is accounted for by family background; 3. Estimate the causal effects of neighborhood environments on the speech patterns of some of our nation's most disadvantaged minority children by exploiting the experimental design of MTO; 4. Use the NLSY97 to establish where in the national speech distribution MTO youth fall as a way of understanding the magnitude of speech impacts and the generalizability of the MTO results; 5. Exploit variation in MTO children's age at baseline to determine how the sensitivity of speech to environmental conditions varies by age (which has been a major question in the socio-linguistics field); 6. Determine whether the speech patterns of male and female youth respond differently to neighborhood social environments, as research in socio-linguistics suggests could be the case and which in turn could potentially explain gender differences in MTO impacts on youth found in previous MTO research; 7. Estimate non-experimentally the association between speech and schooling outcomes in the MTO and NLSY97 data, and whether speech mediates neighborhood effects on schooling; 8. Create versions of the MTO and NLSY97 datasets to be made available for secondary analysis. Successfully addressing these aims would generate new evidence about how and why speech varies across neighborhoods and the potential implications of this speech variation for schooling outcomes, and would also help motivate and guide future research that seeks to identify the causal speech-schooling link.
This proposal requests funding to launch a research program to understand how and why speech patterns vary across neighborhoods, as well as the implications of speech patterns for the schooling outcomes of children growing up in disadvantaged communities. Our research seeks to shed additional light on the out-of-school factors that contribute to disparities in schooling and related life outcomes across neighborhoods, which may help guide the design of new education and social policy interventions to remediate these disparities.
|Rickford, John R; Duncan, Greg J; Gennetian, Lisa A et al. (2015) Neighborhood effects on use of African-American Vernacular English. Proc Natl Acad Sci U S A 112:11817-22|