Predicting Syntax: Processing Dative Constructions in American and Australian Varieties of English

Value of the Information & Strength of the Argument

The talk provides valuable insights into the probabilistic nature of syntax, demonstrating that speakers have implicit knowledge of usage probabilities. The argumentation is solid: the corpus model is rigorously built and validated, and the experimental results support the hypothesis. Bresnan carefully addresses potential objections, such as the equivalence of paraphrases and the role of semantics. The use of naturalistic stimuli and statistical modeling strengthens the claims. However, the talk is a presentation, so some methodological details are abbreviated, and the sample size is modest.

Scientific Rigor, Source Quality, Title Accuracy

The research is grounded in a rich literature on the dative alternation, citing works by Thompson, Hawkins, Collins, Arnold, Wasow, and others. The corpus model is based on the Switchboard corpus, a well-known resource. The experimental design is sound, with careful control and analysis. The title accurately reflects the content. The talk does not include a formal reference list, but the sources are mentioned verbally. The research appears rigorous, though the lack of a published paper for this specific study limits verification.

The title accurately reflects the content: the talk focuses on predicting syntactic choices (dative constructions) in American and Australian English.

8/10

The talk presents original research with a robust methodology: corpus-based modeling validated on unseen data (94% accuracy), experimental design with human subjects, and careful consideration of variables. The speaker is a renowned linguist, and the work is grounded in established literature. However, the presentation is a talk, not a peer-reviewed paper, and some details are omitted.

Markers derived by PSI from the transcript: the creator did not define chapters.

• 02:00Introduction of the speaker by Chris, highlighting her background and contributions.
• 05:00Bresnan outlines the research questions: Can speakers predict syntactic choices? Are reading latencies related to probabilities? Do varieties differ?
• 10:00Explanation of the dative alternation and the assumption of truth-conditional equivalence.
• 15:00Description of the corpus model based on the Switchboard corpus, with variables like discourse accessibility, animacy, and length.
• 20:00Presentation of examples of high, even, and low probability dative constructions.
• 25:00Discussion of the harmonic alignment pattern and its consistency across verb classes.
• 30:00Hypothesis that English speakers know these probabilistic patterns and can use them to predict choices.
• 35:00Description of the experimental design: 30 items, 19 subjects, rating task with 100 points.
• 40:00Results showing a strong correlation between model probabilities and human ratings, with individual variation.
• 45:00Discussion of Australian English data and the similarity of patterns, suggesting shared probabilistic grammar.

• Switchboard Corpus — Used as the source of naturalistic spoken data for the corpus model.
• Bresnan, Cueni, Nikitina, & Baayen (2007) - Predicting the Dative Alternation — The paper describing the corpus model and its variables.
• Collins, P. (1995) - The indirect object construction in English: an informational approach — Cited as a study of the dative alternation in Australian English.

• Bresnan, Cueni, Nikitina, & Baayen (2007) - Predicting the Dative Alternation — The paper presenting the corpus model and its predictions.
• Arnold, Wasow, Losongco, & Ginstrom (2000) - Heaviness vs. newness: The effects of structural complexity and discourse status on constituent ordering — Cited as prior work on factors influencing constituent ordering.

• No discordant sources mentioned — The talk does not present any conflicting sources.