Linguistics Colloquium: Inbal Arnon
Inbal Arnon, The Hebrew University of Jerusalem
Title: The learnability consequences of Zipfian distributions, Ori Lavi-Rotbain & Inbal Arnon
Despite the many differences between them, human languages share certain similarities. These similarities can provide a window onto our shared cognition and the ways in which cognitive biases impact language structure. Here, we focus on the way word frequencies are distributed, one of the striking commonalities between languages. Across languages, word frequencies follow a Zipfian distribution, showing a power law relation between a word's frequency and its rank (Zipf, 1949). Intuitively, this means that languages have relatively few high-frequency words and many low-frequency ones. While studied extensively, little work has explored the learnability consequences of the greater predictability of words in such distributions. Here, we propose that such distributions confer a learnability advantage for various aspects of language learning, and that this advantage may play a part in their propensity in language. We first use corpus analyses to show that child-directed speech is similarly predictable across fifteen different languages. We then experimentally investigate the impact of distribution predictability on children and adults: We present evidence from studies of word segmentation, word learning, and visual statistical learning to illustrate the facilitative effect of skewed distributions on learning and to suggest that learnability pressures play a part in their recurrence in language.
Subscribe to our Telegram channel to get notified about upcoming talks and events