Subject Verb Agreement Error Patterns in Meaningless Sentences: Humans vs. BERT
This work addresses how semantics impacts syntactic processing in language models and humans, with incremental insights into BERT's limitations compared to human cognition.
The study investigated whether meaning interferes with subject-verb agreement in English across various syntactic complexities, comparing BERT-base and humans using both meaningful and nonsensical sentences. It found that both BERT and humans made more errors with nonsensical items, especially with syntactic attractors, and BERT showed a stronger effect of meaningfulness, indicating higher lexical sensitivity.
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.