Combining Language Corpora With Experimental and Computational Approaches for Language Acquisition Research

Abstract

Historically, first language acquisition research was a painstaking process of observation, requiring the laborious hand-coding of children's linguistic productions, followed by the generation of abstract theoretical proposals for how the developmental process unfolds.Recently, the ability to collect large-scale corpora of children's language exposure has revolutionised the field. New techniques enable more precise measurements of children's actual language input, and these corpora constrain computational and cognitive theories of language development, which can then generate predictions about learning behaviour.We describe several instances where corpus, computational, and experimental work have been productively combined to uncover the first language acquisition process and the richness of multimodal properties of the environment, highlighting how these methods can be extended to address related issues in second language research. Finally, we outline some of the difficulties that can be encountered when applying multi-method approaches and show how these difficulties can be obviated.Combining approaches in language acquisition 3 Combining Language Corpora with Experimental and Computational Approaches for Language Acquisition ResearchMultiple methods in language acquisition research are now well-established, although they have not been introduced without difficulty. In this paper, we describe the challenges of combining corpus, experimental, and computational approaches to research in first language acquisition. We discuss the benefits of multi-method approaches, andshow how these allow us to address fundamental questions in first language acquisition, with relevance to related issues in second language learning. Through three examples of successful combination of multiple methods, we illustrate these benefits, and suggest how some of the difficulties of their application may be circumvented for second language acquisition research.Historically, first language acquisition research has been dominated by attempts to describe formal mechanisms that can explain why children acquire the same language structures in the same order, despite great variation in the language environment (Chomsky, 1955;Pinker, 1984). Consequently, much effort in language acquisition research has focussed on determining the universal grammar that described the deep structure (or logical form) of children's linguistic constructions and how this is activated by exposure to a particular language (Chomsky, 1981). Similar arguments have been applied in second language acquisition research in terms of whether we need to posit an innately specified grammar to explain acquisition, or whether there is sufficient positive and negative evidence for the learner to be able to acquire the structure of the language Combining approaches in language acquisition 4 without innate structure (Ellis, 2013;Flynn, Martohardjono, & O'Neil, 1998;Hawkins, 2001;White, 1996).This theoretical approach has been largely unconcerned with combining multiple approaches to i...

Canalization of Language Structure From Environmental Constraints: A Computational Model of Word Learning From Multiple Cues

“…The observation that speed and accuracy of word learning are promoted by multiple cues is consistent with several current accounts of multiple cue integration for various language learning tasks (see Monaghan & Rowland, in press; for a review). For instance, for speech segmentation, Monaghan, White, and Merkx () assessed the acoustic properties of speech to identify multiple prosodic cues that can combine to promote identification of words.…”

Language Learning Research at the Intersection of Experimental, Computational, and Corpus‐Based Approaches

“…Brian MacWhinney argues that further advancement of second language acquisition theory and practice requires a combination of experimental data, a better understanding of how individual differences impact learning, and corpus data that permit the investigation of acquisition patterns. The proposed platform would facilitate this by enabling the collection of substantial amounts of learner data online and by establishing a common protocol on how to share the data—in line with the Child Language Data Exchange System, the central repository for child language data that contributed greatly to our understanding of how children learn language (see Monaghan & Rowland, ). The success of such an approach rests on researchers across the world sharing data and agreeing on common protocols for adding and retrieving data.…”

“…On this basis, they investigated the acquisition of nouns and verbs by adult learners in an artificial language experiment. While artificial language research is occasionally criticized for its limited ecological validity, the use of distributional information from natural language corpora in the artificial language construction mitigates some of this criticism (see also Monaghan & Rowland, ). Another impressive example of multimethod research is Ellis et al.…”

Priming is swell, but it's far from simple

“…The solution is a multi-method approach. As we have argued in Monaghan and Rowland (2017), by gathering evidence from different methods, we can converge on a more holistic understanding of the child's developing representations. Below, we illustrate our argument with two examples.…”