Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Work Smarter...Not Harder: Efficient Minimization of Dependency Length in SOV Languages

Sidharth Ranjan, Titus von der Malsburg

TL;DR

The paper addresses how dependency length minimization operates in SOV languages by proposing a bounded-rationality, least-effort mechanism: place the shortest preverbal constituent adjacent to the main verb to reduce multiple preverbal dependencies. It tests this across seven SOV languages using large-scale UD Treebank data and counterfactual variants, combining corpus analyses with a ranking and regression framework. The findings show corpus sentences align with the least-effort strategy and that this predictor adds explanatory power beyond total dependency length, supporting bounded rationality as a key constraint shaping preverbal ordering and language evolution. The work provides a cross-linguistic mechanistic account of DLM with implications for theories of production, processing, and linguistic evolution under cognitive-resource limits.

Abstract

Dependency length minimization is a universally observed quantitative property of natural languages. However, the extent of dependency length minimization, and the cognitive mechanisms through which the language processor achieves this minimization remain unclear. This research offers mechanistic insights by postulating that moving a short preverbal constituent next to the main verb explains preverbal constituent ordering decisions better than global minimization of dependency length in SOV languages. This approach constitutes a least-effort strategy because it's just one operation but simultaneously reduces the length of all preverbal dependencies linked to the main verb. We corroborate this strategy using large-scale corpus evidence across all seven SOV languages that are prominently represented in the Universal Dependency Treebank. These findings align with the concept of bounded rationality, where decision-making is influenced by 'quick-yet-economical' heuristics rather than exhaustive searches for optimal solutions. Overall, this work sheds light on the role of bounded rationality in linguistic decision-making and language evolution.

Work Smarter...Not Harder: Efficient Minimization of Dependency Length in SOV Languages

TL;DR

The paper addresses how dependency length minimization operates in SOV languages by proposing a bounded-rationality, least-effort mechanism: place the shortest preverbal constituent adjacent to the main verb to reduce multiple preverbal dependencies. It tests this across seven SOV languages using large-scale UD Treebank data and counterfactual variants, combining corpus analyses with a ranking and regression framework. The findings show corpus sentences align with the least-effort strategy and that this predictor adds explanatory power beyond total dependency length, supporting bounded rationality as a key constraint shaping preverbal ordering and language evolution. The work provides a cross-linguistic mechanistic account of DLM with implications for theories of production, processing, and linguistic evolution under cognitive-resource limits.

Abstract

Dependency length minimization is a universally observed quantitative property of natural languages. However, the extent of dependency length minimization, and the cognitive mechanisms through which the language processor achieves this minimization remain unclear. This research offers mechanistic insights by postulating that moving a short preverbal constituent next to the main verb explains preverbal constituent ordering decisions better than global minimization of dependency length in SOV languages. This approach constitutes a least-effort strategy because it's just one operation but simultaneously reduces the length of all preverbal dependencies linked to the main verb. We corroborate this strategy using large-scale corpus evidence across all seven SOV languages that are prominently represented in the Universal Dependency Treebank. These findings align with the concept of bounded rationality, where decision-making is influenced by 'quick-yet-economical' heuristics rather than exhaustive searches for optimal solutions. Overall, this work sheds light on the role of bounded rationality in linguistic decision-making and language evolution.
Paper Structure (11 sections, 2 equations, 4 figures, 2 tables)

This paper contains 11 sections, 2 equations, 4 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Data and Methods
  3. Results
  4. Constituent Length Analysis
  5. Counterfactual Analysis
  6. Corpus Prediction Task
  7. Ranking Model
  8. Regression Analysis
  9. Prediction Accuracy
  10. Discussion
  11. Acknowledgments

Figures (4)

  • Figure 1: Preverbal constituent ordering problem in a language with subject-object-verb (SOV) word order; Constituent (C$_{i}$) is defined as a word or a group of words that functions as a single unit within a syntactic configuration
  • Figure 2: Average constituent length of preverbal constituents for corpus sentences with 2 to 5 constituents separately
  • Figure 3: Preverbal constituent ordering in Hindi (SOV); Only main verb dependencies are depicted; Total dependency length (DL) and Constituent's DL are indicated above each arc; Image reproduced from RanjanMalsburg2023CogSci
  • Figure 4: Average total dependency length (i.e., total dependency length normalized by the number of words in a sentence) for different constituent orderings and for different numbers of preverbal constituents