In the domain of Linguistics, the categories of intransitive verbs, namely the unaccusativity, is a long-debated topic. Unaccusativity suggests that intransitive verbs can be divided into unergative and unaccusative verbs, based on their subjects' similarity to the subjects of transitive verbs or the objects of transitive verbs. Previous research has discussed how the meaning of verbs can decide the unaccusativity of intransitive verbs, but the meanings of verbs alone still cannot predict the unaccusativity of intransitive verbs cross-linguistically. Moreover, while the sentential environment can have an impact on the categories of intransitive verbs, previous studies did not investigate how the environment plays a role in the categories. This paper examines this issue from child language acquisition. I select a few sentential environments in the children's corpus of Mandarin and conduct a qualitative analysis that suggests that these sentence environments indeed possess the properties of either category. In a child acquisition experiment, I show that when the category of verbal meanings and sentential environments align, the categorization of verbs is the most obvious and efficient. I introduce the concept of ‘compatibility' to describe this relationship between verb meaning and the sentential environment. These results suggest that speakers can infer the unaccusativity of verbs from a variety of sentence environments in language that may not be directly linked to the concept of unaccusativity, and the concept of ‘compatibility' in language environment is a crucial factor in the categories/categorization of unaccusativity.

Intransitive verbs in human language can be subcategorized into at least two linguistic categories, unaccusative and unergative verbs. The categorization of these verbs has been a subject of debate, with Projectional approaches emphasizing the role of verb semantics and Constructional approaches highlighting the importance of the sentence context. We utilize a word2vec model to capture the environmental influence on verb semantics, providing evidence that supports both approaches. Our results demonstrate that the categorization of Unaccusativity for new verbs can be influenced by the sentence context in which they appear, and the frequency of verbs in specific contexts also plays a significant role. Additionally, through a child language acquisition experiment, we show that the sentence environment has a significant impact on the categorization of Unaccusativity when the semantics of new verbs are provided. These findings suggest that both approaches have merit, highlighting the psychological universality of Unaccusativity across languages.