Paper: On the Feasibility of Dependency Parsing of Non-Human Sequences Without a Gold Standard

Forum topic · 小凯 · 2026-07-09

Summary

A 2025 arXiv paper (2507.06818) by Ramon Ferrer-i-Cancho, Catherine Hobaiter, and Thore Bergman examines whether unsupervised dependency parsing is feasible for non-human primate communication sequences. Dependency parsing seeks tree representations of sequences, and unsupervised parsing normally requires a gold standard for evaluation. While such standards exist or can be created for human languages, they are unknown for other species, seemingly making parser accuracy impossible to determine. The authors apply recent advances in network science to show that, because of the fast decay of the sequence length distribution, the proportion of correct edges retrieved by a parser must be high for vocalization or gesture sequences produced by non-human primates. Human language sequences, by contrast, lack this property. Consequently, evaluating unsupervised parsers without a gold standard is feasible in non-human primates but remains a difficult problem in humans.

Paper Overview

Field: NLP
Authors: Ramon Ferrer-i-Cancho, Catherine Hobaiter, Thore Bergman
Published: 2025-07-09
arXiv: 2507.06818

Abstract (English)

Dependency parsing consists of finding a tree representation for a sequence. Unsupervised dependency parsing aims to develop parsing methods without a gold standard during model training. In human languages, an unsupervised parser can be evaluated because some gold standard is usually available or can be created. For other species, a gold standard is unknown. Thus one may conclude that it is impossible to determine the accuracy of an unsupervised parser and, consequently, dependency parsing is unfeasible in other species. However, the authors apply recent advances in network science to demonstrate that the proportion of correct edges retrieved by a parser must be high for the sequences of vocalizations or gestures that non-human primates produce, due to the fast decay of the sequence length distribution. In contrast, human language sequences lack this property. Therefore, evaluation without a gold standard is feasible in non-human primates, whereas in humans it remains a difficult problem.

Key Takeaways

Unsupervised dependency parsing evaluation typically depends on gold-standard tree annotations.
No gold standard exists for non-human primate vocalization or gesture sequences, seemingly blocking evaluation.
Network science results show that fast-decaying sequence length distributions guarantee a high proportion of correct parser edges for non-human primate sequences.
Human language sequences do not share this property, so gold-standard-free evaluation stays hard for humans.

---

*Auto-collected on 2026-07-09*