The computational BIA+ model (Dijkstra & Van Heuven, 2002) has provided a useful account for bilingual word recognition, while the verbal (pre-quantitative) RHM (Kroll & Stewart, 1994) has often served as a reference framework for bilingual word production and translation. According to Brysbaert and Duyck (2010), a strong need is felt for a unified implemented account of bilingual word comprehension, lexical-semantic processing, and word production. With this goal in mind, we built a localist-connectionist model, called Multilink, which integrates basic assumptions of both BIA+ and RHM. It simulates the recognition and production of cognates (form-similar translation equivalents) and non-cognates of different lengths and frequencies in tasks like monolingual and bilingual lexical decision, word naming, and word translation production. It also considers effects of lexical similarity, cognate status, relative L2-proficiency, and translation direction. Model-to-model comparisons show that Multilink provides higher correlations with empirical data than both IA and BIA+ models.

Like the BIA model (Dijkstra & van Heuven, 1998; van Heuven, Dijkstra & Grainger, 1998) and the BIA+ model (Dijkstra & van Heuven, 2002), the Multilink model is a symbolic, localist-connectionist, interactive model for lexical processing in the visual domain. In our view, the symbolic nature of Multilink makes it attractive and easily interpretable, even in relation to brain activity (Page, 2000, p. 501; 2017). Symbolic localist-connectionist models have a long tradition and have been applied to many different areas of cognitive research (e.g., Grainger & Jacobs, 1998). As a consequence, a lot is known about their properties and limitations (e.g., Bowers, 2009). These models can also easily be organized hierarchically in a cognitive functional way, and they have a reasonable degree of flexibility while still being falsifiable. Thus, despite the availability of other sophisticated frameworks for modeling language processes, a lot can still be gained from localist models.

Much recent work on language and cognition has examined the psychological status of collocations/formulas/multi-word expressions as mentally stored units. These studies have used a variety of statistical metrics to quantify the degree of strength or association of these sequences, and then they have correlated these strengths with particular behavioral effects that evidence mental storage. However, the relationship between intonational prosody and storage of collocations has received little attention. Through a corpus-based approach, this study examines the hypothesis that boundaries between successive intonation units avoid splitting word bigrams that exhibit high statistical association, with such high association taken to be an index of mental storage of these bigrams. Conversely, bigrams exhibiting lower statistical association ought to be more likely to be split by intonation unit boundaries under this hypothesis.

No abstract

We describe a new algorithm for the extraction of formulaic language from corpora. Entitled MERGE (Multi-word Expressions from the Recursive Grouping of Elements), it iteratively combines adjacent bigrams into progressively longer sequences based on lexical association strengths. We then provide empirical evidence for this approach via two case studies. First, we compare the performance of MERGE to that of another algorithm by examining the outputs of the approaches compared with manually annotated formulaic sequences from the spoken component of the British National Corpus. Second, we employ two child language corpora to examine whether MERGE can predict the formulas that the children learn based on caregiver input. Ultimately, we show that MERGE indeed performs well, offering a powerful approach for the extraction of formulas.