Recurrent neural networks (RNNs) have achieved impressive results in a variety of linguistic processing tasks, suggesting that they can induce non-trivial properties of language. We investigate here to what extent RNNs learn to track abstract hierarchical syntactic structure. We test whether RNNs trained with a generic language modeling objective in four languages (Italian, English, Hebrew, Rus...

Automatically judging sentences for their grammaticality is potentially useful for several purposes — evaluating language technology systems, assessing language competence of second or foreign language learners, and so on. Previous work has examined parser ‘byproducts’, in particular parse probabilities, to distinguish grammatical sentences from ungrammatical ones. The aim of the present paper ...

Language acquisition is often said to require a massive innate body of language speciic knowledge in order to overcome the poverty of stimulus. In this picture, language learning merely implies setting a number of parameters in an internal Universal Grammar. But is the primary linguistic evidence really so poor that it warrants such an extreme nativism? Is there no room for a more empiricist or...

According to the standard definition, linguistic competence, the object of study of generative grammar, is knowledge of language. As it is used in generative linguistics, however, knowledge of language is also a technical term, referring to some system of rules that determines the set of sentences in a language. As such, any linguistic knowledge that does not bear on the grammaticality or lingu...

1 I aim to discuss and evaluate my acquisition of the grammatical systems of foreign languages; focusing primarily on Finnish, though with reference to German. Having done a short three week course for total beginners in Finnish and having studied German to undergraduate level, I have achieved a very different level of competence in the languages. These two entirely separate learning experience...

Researchers have long been puzzled by children's variable omission of grammatical morphemes, often attributing this to a lack of semantic or syntactic competence. Recent studies suggest that some of this variability may be due to phonological constraints. This paper explored this issue further by conducting a longitudinal study of five English-speaking one- to two-year-olds' acquisition of arti...

Abstract In this work, we present a phenomenon-oriented comparative analysis of the two dominant approaches in English Resource Semantic (ERS) parsing: classic, knowledge-intensive and neural, data-intensive models. To reflect state-of-the-art neural NLP technologies, factorization-based parser is introduced that can produce Elementary Dependency Structures much more accurately than previous da...

In his paper “Do young children have adult syntactic competence?” Tomasello (Cognition 74 (2000) 209) interprets young children’s conservatism in language production as evidence that early language use, and verb use in particular, are based entirely on concrete lexical representations, showing no evidence of abstract syntactic categories such as “verb” or “transitive sentence”. In this reply, I...

Half a century ago, Noam Chomsky introduced the field of linguistics to new mathematical tools drawn largely from recursive function theory. These were exciting tools that imparted mathematical precision, perhaps for the first time in the history of Linguistics, to the enterprise of grammar construction. The cornerstone of Chomsky’s new theoretical edifice, was the grammatical transformation, a...

The idea of a perfectly competent but resource limited language user is the basis of many models of sentence comprehension. It is widely assumed that linguistic competence is a) uniform; b) generative; c) autonomous; d) automatic and e) constant. It is also believed that the free expression of these properties is frustrated by limits in the availability of computational resources. However, no f...