Subject Verb Agreement Error Patterns in Meaningless Sentences: Humans vs. BERT

TL;DR

This study compares how humans and BERT handle subject-verb agreement in meaningful versus meaningless sentences, revealing that semantics influence both, but more strongly affect BERT's performance.

Contribution

It demonstrates that semantics interfere with syntactic agreement tasks in both humans and BERT, with BERT showing greater lexical sensitivity than humans.

Findings

Both humans and BERT perform worse on nonsensical sentences.

Semantic interference is stronger in BERT than in humans.

Attractors increase SVA errors for both humans and BERT.

Abstract

Both humans and neural language models are able to perform subject-verb number agreement (SVA). In principle, semantics shouldn't interfere with this task, which only requires syntactic knowledge. In this work we test whether meaning interferes with this type of agreement in English in syntactic structures of various complexities. To do so, we generate both semantically well-formed and nonsensical items. We compare the performance of BERT-base to that of humans, obtained with a psycholinguistic online crowdsourcing experiment. We find that BERT and humans are both sensitive to our semantic manipulation: They fail more often when presented with nonsensical items, especially when their syntactic structure features an attractor (a noun phrase between the subject and the verb that has not the same number as the subject). We also find that the effect of meaningfulness on SVA errors is stronger for BERT than for humans, showing higher lexical sensitivity of the former on this task.