Word learning and the acquisition of syntactic-semantic overhypotheses

By Jon Gauthier, Roger Levy, Josh Tenenbaum. 2018

11 citations recorded by the source corpus.

Children learning their first language face multiple problems of induction: how to learn the meanings of words, and how to build meaningful phrases from those words according to syntactic rules. We consider how children might solve these problems efficiently by solving them jointly, via a computational model that learns the syntax and semantics of multi-word utterances in a grounded reference game. We select a well-studied empirical case in which children are aware of patterns linking the syntactic and semantic properties of words --- that the properties picked out by base nouns tend to be related to shape, while prenominal adjectives tend to refer to other properties such as color. We show that children applying such inductive biases are accurately reflecting the statistics of child-directed speech, and that inducing similar biases in our computational model captures children's behavior in a classic adjective learning experiment. Our model incorporating such biases also demonstrates a clear data efficiency in learning, relative to a baseline model that learns without forming syntax-sensitive overhypotheses of word meaning. Thus solving a more complex joint inference problem may make the full problem of language acquisition easier, not harder.

Subfields: Artificial Intelligence, Cultural Studies, Developmental and Educational Psychology, Cognitive Science & Psychology, Linguistics

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Josh Tenenbaum
2. Chuang Gan
3. Jiajun Wu
4. Song-Chun Zhu
5. Pushmeet Kohli
6. Jiayuan Mao
7. Yining Hong
8. Siyuan Huang

In Cultural Studies

1. Roger Levy
2. Song-Chun Zhu
3. Josh Tenenbaum
4. Tomer D. Ullman
5. Brenden M. Lake
6. Laura Schulz
7. Yangqing Jia
8. Michael D. Lee

In Developmental and Educational Psychology

1. Song-Chun Zhu
2. Jiajun Wu
3. Pushmeet Kohli
4. Roger Levy
5. Josh Tenenbaum
6. Tomer D. Ullman
7. T. Griffiths
8. Noah D. Goodman

In Cognitive Science & Psychology

1. Roger Levy
2. Chuang Gan
3. Song-Chun Zhu
4. Josh Tenenbaum
5. Jiajun Wu
6. Tomer D. Ullman
7. Brenden M. Lake
8. Michael D. Lee

In Linguistics

1. Roger Levy
2. Song-Chun Zhu
3. Josh Tenenbaum

Citations in the corpus

• Grammar-Based Grounded Lexicon Learning (2022, background): “For this reason, previous approaches to lexicon learning have either assumed an expert-annotated set of lexical entries [48] or only attempted to learn at very small scales [16].”
• The Neuro-Symbolic Concept Learner: Interpreting Scenes, Words, and Sentences From Natural Supervision (2019, background): “We leave the automatic discovery of concept words as a future work (Gauthier et al., 2018).”
• A Competence-aware Curriculum for Visual Concepts Learning via Question Answering (2020, background): “Humans excel at learning visual concepts and their compositions in a questionanswering manner [16,10,18], which requires a joint understanding of vision and language.”
• Visually Grounded Neural Syntax Acquisition (2019, unknown): “Automatically acquiring such inductive biases from data remains challenging (Kemp et al., 2006; Gauthier et al., 2018).”