There is an ongoing debate of whether neural network can grasp the quasi-regularities in languages like humans. In a typical quasi-regularity task, English past tense inflections, the neural network model has long been criticized that it learns only to generalize the most frequent pattern, but not the regular pattern, thus can not learn the abstract categories of regular and irregular and is dissimilar to human performance. In this work, we train a set of transformer models with different settings to examine their behavior on this task. The models achieved high accuracy on unseen regular verbs and some accuracy on unseen irregular verbs. The models’ performance on the regulars are heavily affected by type frequency and ratio but not token frequency and ratio, and vice versa for the irregulars. The different behaviors on the regulars and irregulars suggest that the models have some degree of symbolic learning on the regularity of the verbs. In addition, the models are weakly correlated with human behavior on nonce verbs. Although the transformer model exhibits some level of learning on the abstract category of verb regularity, its performance does not fit human data well suggesting that it might not be a good cognitive model.

Case is an abstract grammatical feature that indicates argument relationship in a sentence. In English, cases are expressed on pronouns, as nominative case (e.g. I, he), accusative case (e.g. me, him) and genitive case (e.g. my, his). Children correctly use cased pronouns at a very young age. How do they acquire abstract case in the first place, when different cases are not associated with different meanings? This paper proposes that the distributional patterns in parents’ input could be used to distinguish grammatical cases in English.