Lexical Tone as a Cue in Statistical Word Learning from Bilingual Input

Abstract

Learners can track word-referent co-occurrences across individually-ambiguous naming events to form correct word-referent mappings, termed statistical word learning (SWL). Prior research largely focuses on learning from a single language input, where a referent co-occurs with a single word (1:1 mapping). Here, we tested adults’ SWL from a simulated bilingual environment, where one referent co-occurred with two words (2:1 mapping) and the two words were either differentiated by a linguistic cue (Mandarin lexical tones, Cued condition) or not (Uncued condition). Results showed that in the Cued condition, Chinese-English bilinguals (N = 38) outperformed Spanish-English bilinguals (N = 56) and English monolinguals (N = 55), while Spanish-English bilinguals and English monolinguals performed similarly. The three groups did not differ in the Uncued condition. Self-reported learning confidence and strategies showed limited conscious awareness of learning. Results demonstrate that familiarity with a linguistic cue boosts overall statistical word learning from bilingual input.

Cross-Situational Word Learning in Children and Adults: The Case of Lexical Overlap

“…A limited set of studies have examined how adult learners deal with multiple ambiguous learning events that present lexical overlap. This work has demonstrated that adults are successful at cross-situational word learning when a single word consistently co-occurs with two referents (one-to-two structure; Poepsel & Weiss, 2016;Yurovsky et al, 2013) or, most relevant to our current study, when two words consistently co-occur with the same referent (two-to-one structure; Benitez et al, 2016;Chan & Monaghan, 2019;Ichinco et al, 2009;Kachergis et al, 2012;Li & Benitez, 2023). However, adults show an advantage in performance for one-to-one compared to two-to-one structure, suggesting that learning from two-to-one structure can be difficult (but see Chan & Monaghan, 2019 for the opposite pattern of results).…”

“…Might pragmatic cues also benefit cross-situational word learning of overlapping structure? Only a handful of studies have examined this question, with mixed findings (Benitez et al, 2016;Crespo & Kaushanskaya, 2021;Chan & Monaghan, 2019;Li & Benitez, 2023;Poepsel & Weiss, 2014). Language cues may be particularly informative.…”

“…Language cues may be particularly informative. In a recent study testing adults, we examined the role of lexical tone as a cue signaling that the two words sharing a referent stem from different languages (Li & Benitez, 2023). Results showed that the cue of lexical tone provided a boost in cross-situational word learning of lexical overlap, but only for adults with experience with lexical tone (Chinese-English bilinguals).…”

“…For example, bilingual children, who encounter lexical overlap from both languages frequently, have been found to be less likely to hold a mutual exclusivity bias than monolingual children (Bialystok et al, 2010;Byers-Heinlein & Werker, 2009;Davidson et al, 1997;Houston-Price et al, 2010;Kalashnikova et al, 2015). In cross-situational word learning, however, there are mixed findings with respect to whether bilingualism impacts learning (Benitez et al, 2016;Crespo & Kaushanskaya, 2021;Escudero et al, 2016;Li & Benitez, 2023;Poepsel & Weiss, 2016;Chan & Monaghan, 2019). One study found an advantage for bilingual over monolingual adults in one-to-two (one label, two referents)…”

“…but not one-to-one learning (Poepsel & Weiss, 2016); two studies found an advantage for bilingual over monolingual adults in two-to-one learning, but only when a language cue differentiated the two labels for a referent (Benitez et al, 2016;Li & Benitez, 2023); and yet another study found no differences between bilingual and monolingual adults in one-to-one or two-to-one learning, but found that bilinguals had a faster learning rate than monolinguals (Chan & Monaghan, 2019). These findings highlight the need to measure learners' language experience with lexical overlap when assessing cross-situational word learning from overlapping structure.…”