Jennifer K. Spenader

There are a number of proposals on the Aging Lab site. Below are proposal for testing young adults.

"Proposal: Do offline preferences for antecedents of Dutch demonstrative pronouns show up in online measurements?

SHORT DESCRIPTION: EXPERIMENTAL PROJECT: The student will carry out an experiment testing the differences in interpretation between personal pronouns like 'hij' and demonstrative pronouns like 'die' using an online measure: self-paced reading.
LONGER DESCRIPTION:
Previous research has shown that for Dutch (Kaiser 2011), personal pronouns prefer subject antecedents while demonstrative pronousn prefer object antecedents. When do you have to use cognitive resources to interpret an ambiguous pronoun? Consider the following sentences:
(1) De brandweerman kneep de bokser speels. Hij/Die . . .

In sentence (1), the pronoun 'Hij' generally gets interpreted as the fireman, while the demonstrative pronoun 'Die' generally gets interpreted as the boxer. This is been shown with offline completion studies and with visual-world eye-tracking. But do we see the same effects in self-paced reading? A recent paper from Bosch and Umbach found less clear cut results for German.

In the current project we will use existing materials from previous experiments with Dutch (Kaiser 2011), but with self-paced reading as the measure of online processing.

KEY REFERENCES:

Kaiser (2011). Salience and contrast effects in reference resolution: The interpretation of Dutch pronouns and demonstratives, Language and Cognitive Processes, 26, 1587-1624.

"Proposal: What contextual factors influence pronoun interpretation.

This project will have Margreet Vogelzang as the external advisor (and Jennifer Spenader as the internal advisor). Contact Margreet if you are interested (margreet.vogelzang@rug.nl). Overall project description: Dutch children make errors in their comprehension of object pronouns until the age of 6 (e.g., Koster, 1993; Philip & Coopmans, 1996). For them, the sentence ``Bert is washing him'' can also mean that Bert is washing himself. Italian children, in contrast, do not make such errors in their language (McKee, 1992). This project investigates why Dutch and Italian children differ. It compares Dutch and Italian children and adults on their comprehension of pronouns in discourse, using various experimental methods (for example, eyetracking) as well as cognitive modeling to computationally simulate the process of pronoun resolution. Ultimately, this project hopes to shed more light on the issue whether and when the listeners interpretation of a pronoun is guided by discourse information and whether and when it is guided by the speaker's choices.

Production study In order to find out more about how the production and comprehension of pronouns relate, we are looking for a student to perform a production study eliciting the use of subject (and object) pronouns in Dutch adults. In an experimental set-up where one participant has to communicate a particular series of events with two animate referents of the same (vs. a different) gender to someone else, it is investigated whether participants prefer to use pronouns for prominent referents and therefore sometimes produce ambiguous sentences, or whether they prefer to use more explicit full noun phrases to make the intended meaning clear to the listener. The student for this project is encouraged to discuss and further develop the experimental setup together with the daily advisor. The student will gain experience in the testing of human subjects, and in the complete process of developing an experiment. The student(s) will be encouraged to present the results at a workshop or conference. We will also aim at publishing the results with the student(s) as co-authors.

Corpus study To find out more about the usage of pronouns in non-experimental settings, we are looking for a student to perform a corpus study investigating the linguistic contexts in which pronouns such as the Dutch hem `him', or the combination of the pronouns hij `he' and hem `him' (or the equivalent forms in English or Italian) are produced. Do writers avoid using an object pronoun if there are more referents of the same gender (compared to different gender)? That is, do they try to avoid being misunderstood by the listener? Or do they use pronouns to signal referents that are prominent in the discourse, at the risk of producing ambiguous sentences? The student for this project will have an active role in determining the research questions to be investigated. Furthermore, the student will gain experience with corpus research.