The CHILDES project: tools for analyzing talk

Simulating the cross-linguistic pattern of Optional Infinitive errors in children’s declaratives and Wh- questions

“…However, it is also clear that the German child, Leo, and the Spanish child, Juan, produce few Wh-questions of any kind at T1. In order to obtain a more reliable estimate of the proportion of OI errors in Wh-questions in German and Spanish during the early stages, we therefore analysed the data from the 6 typically developing Germanspeaking children in the Szagun corpus (Szagun, 2001) and 5 Spanish-speaking children from the Aguirre, Irene, Marrero/Albalá and Ornat corpora (López Ornat, 1994), all of which are available in the CHILDES database (MacWhinney, 2000). The results of this analysis are presented in Table 3.…”

Bootstrapping in a language of thought: A formal model of numerical concept learning

“…The likelihood uses a second free parameter, α, which controls the degree to which the learner is penalized for data which does not agree with their hypothesis. To create data for the learning model, we simulated noisy pairing of words and sets of objects, where the word frequencies approximate the naturalistic word probabilities in child-directed speech from CHILDES (MacWhinney, 2000). We used all English transcripts with children aged between 20 and 40 months to compute these probabilities.…”

The company that words keep: comparing the statistical structure of child- versus adult-directed language

“…Corpora and free associates A corpus taken from the American section of the CHILDES database (MacWhinney, 2000), representing transcripts of caregiver speech directed to children aged 1 ; 0 to 5; 0 (provided by Riordan & Jones, 2007 ; see also Riordan & Jones, 2011), were divided into four consecutive twelve-month periods. These four corpora consisted of approximately 500,000 words each, consisting of only adult speech.…”