parsing-as-language-modeling-choe-charniak-2016
IN premise — summaries/2026/08/24/wiki-Natural_language_processing-chunk-3.md
Created 2026-08-24T17:11:19+00:00
Choe & Charniak (2016) proposed parsing as a language modeling task, and Vinyals et al. (2015) framed grammar induction as sequence-to-sequence learning, representing the shift from hand-built parsers to neural approaches.
Summary
This marks the turning point where researchers stopped hand-coding grammatical rules and instead let neural networks learn to parse sentences from data, treating grammar the same way a language model learns to predict words. It matters because it reframes grammar as a pattern to be discovered rather than a structure to be engineered, and nearly all modern parsing and grammar-learning work builds on that shift.