# Unsupervised Parsing via Constituency Tests

arXiv: https://arxiv.org/abs/2010.03146 · HTML: https://mlsanity.com/paper/2010.03146/

## Key facts

- Cited by 23 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 7 Oct 2020.

## Built on by (most cited first, 23 of 23)

- [LM-Critic: Language Models for Unsupervised Grammatical Error Correction](https://mlsanity.com/paper/2109.06822/) (2021)
- [PCFGs Can Do Better: Inducing Probabilistic Context-Free Grammars with Many Symbols](https://mlsanity.com/paper/2104.13727/) (2021)
- [Neural Bi-Lexicalized PCFG Induction](https://mlsanity.com/paper/2105.15021/) (2021)
- [Dynamic Programming in Rank Space: Scaling Structured Inference with Low-Rank HMMs and PCFGs](https://mlsanity.com/paper/2205.00484/) (2022)
- [Contextual Distortion Reveals Constituency: Masked Language Models are Implicit Parsers](https://mlsanity.com/paper/2306.00645/) (2023)
- [Improved Latent Tree Induction with Distant Supervision via Span Constraints](https://mlsanity.com/paper/2109.05112/) (2021)
- [Ensemble Distillation for Unsupervised Constituency Parsing](https://mlsanity.com/paper/2310.01717/) (2023)
- [Learning Syntax from Naturally-Occurring Bracketings](https://mlsanity.com/paper/2104.13933/) (2021)
- [Syntactic Substitutability as Unsupervised Dependency Syntax](https://mlsanity.com/paper/2211.16031/) (2022)
- [Co-training an Unsupervised Constituency Parser with Weak Supervision](https://mlsanity.com/paper/2110.02283/) (2021)
- [ContextRef: Evaluating Referenceless Metrics For Image Description Generation](https://mlsanity.com/paper/2309.11710/) (2023)
- [Simple Hardware-Efficient PCFGs with Independent Left and Right Productions](https://mlsanity.com/paper/2310.14997/) (2023)
- [Re-evaluating the Need for Multimodal Signals in Unsupervised Grammar Induction](https://mlsanity.com/paper/2212.10564/) (2022)
- [Constituency Parsing using LLMs](https://mlsanity.com/paper/2310.19462/) (2023)
- [Unsupervised Parsing by Searching for Frequent Word Sequences among Sentences with Equivalent Predicate-Argument Structures](https://mlsanity.com/paper/2404.12059/) (2024)
- [Bridging Pre-trained Language Models and Hand-crafted Features for Unsupervised POS Tagging](https://mlsanity.com/paper/2203.10315/) (2022)
- [Improving Unsupervised Constituency Parsing via Maximizing Semantic Information](https://mlsanity.com/paper/2410.02558/) (2024)
- [Implicit Representations of Grammaticality in Language Models](https://mlsanity.com/paper/2605.05197/) (2026)
- [Focused Contrastive Training for Test-based Constituency Analysis](https://mlsanity.com/paper/2109.15159/) (2021)
- [Tree-Averaging Algorithms for Ensemble-Based Unsupervised Discontinuous Constituency Parsing](https://mlsanity.com/paper/2403.00143/) (2024)
- [On Eliciting Syntax from Language Models via Hashing](https://mlsanity.com/paper/2410.04074/) (2024)
- [Understanding Syntactic Generalization in Structure-inducing Language Models](https://mlsanity.com/paper/2508.07969/) (2025)
- [Holographic Neural PCFG for Unsupervised Parsing](https://mlsanity.com/paper/2607.08063/) (2026)

## Abstract (arXiv)

We propose a method for unsupervised parsing based on the linguistic notion of a constituency test. One type of constituency test involves modifying the sentence via some transformation (e.g. replacing the span with a pronoun) and then judging the result (e.g. checking if it is grammatical). Motivated by this idea, we design an unsupervised parser by specifying a set of transformations and using an unsupervised neural acceptability model to make grammaticality decisions. To produce a tree given a sentence, we score each span by aggregating its constituency test judgments, and we choose the binary tree with the highest total score. While this approach already achieves performance in the range of current methods, we further improve accuracy by fine-tuning the grammaticality model through a refinement procedure, where we alternate between improving the estimated trees and improving the grammaticality model. The refined model achieves 62.8 F1 on the Penn Treebank test set, an absolute improvement of 7.6 points over the previous best published result.