Revisiting the poverty of the stimulus: hierarchical generalization without a hierarchical bias in recurrent neural networks

TL;DR

This study investigates whether recurrent neural networks can learn hierarchical syntactic rules without explicit hierarchical biases, emphasizing the role of hierarchical cues like subject-verb agreement in language acquisition.

Contribution

It demonstrates that certain RNN architectures can acquire hierarchical syntactic rules using cues present in the language input, challenging the need for explicit hierarchical biases.

Findings

Some RNN architectures learn hierarchical rules

Hierarchical cues like subject-verb agreement improve learning

RNNs can simulate aspects of language acquisition

Abstract

Syntactic rules in natural language typically need to make reference to hierarchical sentence structure. However, the simple examples that language learners receive are often equally compatible with linear rules. Children consistently ignore these linear explanations and settle instead on the correct hierarchical one. This fact has motivated the proposal that the learner's hypothesis space is constrained to include only hierarchical rules. We examine this proposal using recurrent neural networks (RNNs), which are not constrained in such a way. We simulate the acquisition of question formation, a hierarchical transformation, in a fragment of English. We find that some RNN architectures tend to learn the hierarchical rule, suggesting that hierarchical cues within the language, combined with the implicit architectural biases inherent in certain RNNs, may be sufficient to induce hierarchical generalizations. The likelihood of acquiring the hierarchical generalization increased when the language included an additional cue to hierarchy in the form of subject-verb agreement, underscoring the role of cues to hierarchy in the learner's input.