Cultural Input and Conceptual Development

This line of research focuses on how cultural input, specifically linguistic experience, shapes conceptual development. For example, generic language (e.g., “ducks lay eggs,” “girls hate math”) shapes beliefs about category structure. I study the mechanisms through which generics have these effects. In my research with Dr. Rhodes and Dr. Sarah-Jane Leslie at Princeton University, we found evidence in support of a novel hypothesis—that the form of generic statements is a crucial aspect of how they shape beliefs about categories, beyond the content of the properties they assert. That is, referring to a kind using generic statements (e.g., “Zarpies have striped hair”) communicates the belief that the kind (e.g., “Zarpies”) is meaningful and informative, and introduces this belief as common knowledge. This proposal makes a striking prediction, which is that hearing generic language could shape category beliefs even if the information they convey later turns out to be wrong (e.g., a more knowledgeable speaker says, “that’s not true about Zarpies!”). Both statements confirm the assumption that Zarpies have properties in common. Rather, in order to undermine the belief that the kind is meaningful, one would need to correct the generic scope of the statements (e.g., “no, just this Zarpie has striped hair”). In support of this proposal, correcting the scope of generic statements, to specific individuals, led children to think of novel social and animal categories as less meaningful and homogenous than correcting their content (Foster-Hanson, Leslie, & Rhodes, submitted).

More recently, we've found that limiting the scope of generic statements even shapes beliefs about gender categories—categories that young children already think of as meaningful (here's a PDF of my recent poster on this work presented at the 2019 meeting of the Cognitive Development Society). This finding offers exciting implications for how to correct some of the generic statements children hear in their daily lives, in order to limit social stereotyping. In this vein, I'm working with Dr. Rhodes and my mentee Josie Benitez (who is currently pursuing a PhD in Dr. Rhodes’ lab) to replicate previous work measuring parents’ use of generics when talking to their children about gender categories—and children’s subsequent beliefs about gender—using updated materials and a broader set of measures. In the next phase of this project, we'll develop an intervention strategy to help parents talk about gender in ways that lessen children’s gender stereotyping.

A particularly exciting aspect of this project is that we're collecting data online through the Princeton and NYU Discoveries in Action (PANDA) web platform, which lets kids and parents participate in studies from home while recording data through their computer's webcam. This type of unmoderated remote research helps to increase the diversity of research samples, by removing barriers to participation like geographic location and physical obstacles to coming in to the lab. This platform will also enable us to record longitudinal data, so that we can measure how changes in parental speech shape beliefs across childhood. We describe an accessible approach to implementing unmoderated remote research in developmental science, and discuss some of this method's  challenges and its potential for advancing the field, in a recent paper (Rhodes, M., Rizzo, M., Foster-Hanson, E., Moty, K., Leshin, R., Wang, M. M., Benitez, J., & Ocampo, J. D., in press). We also created an accompanying website, discoveriesonline.org, to help other researchers get started with unmoderated remote research.

In previous work, I've also studied how linguistic input shapes children’s attitudes and behavior in their daily lives. For example, in my work with Dr. Rhodes, Dr. Andrei Cimpian, and my mentee Rachel Leshin (who is currently pursuing a PhD in Dr. Rhodes’ lab), we found that children who were asked to “be helpers,” and then encountered setbacks while trying to help, had lower self-evaluations and helped less on subsequent tasks, than children asked to “help” (Foster-Hanson, Cimpian, Leshin, & Rhodes, 2018).