How Individual Differences Affect Learning of Translation‐Ambiguous Vocabulary

Abstract

This study examined interactions of word and learner characteristics during foreign vocabulary learning, focusing on translation ambiguity and individual differences in cognitive resources and linguistic background (language proficiency, multilingual experience). Fifty-three native Hebrew speakers and Russian-Hebrew multilinguals learned the phonological form of target Arabic words along with their Hebrew translations and definitions. The mapping could be translation ambiguous, with a single Hebrew word translated into two Arabic words (one-to-many) or translation unambiguous (oneto-one mapping). Results from translation production and meaning recognition tests revealed that translation-ambiguous words were more difficult to learn than translationunambiguous words. This disadvantage did not dissipate with time, and learners' phonological short-term memory was associated with increased translation ambiguity costs. Learners' proficiency in the language through which learning took place (Hebrew), but not degree of multilingualism, modulated learning. Findings underscore the importance of item and learner interactions, clarifying the multilingualism effect in novel language learning. active competition between the different alternative translations may hinder performance, which is in line with assumptions of competitive inhibitory interactions between connected representations in interactive activation models (Doherty, 2004;McClelland & Rumelhart, 1981). Second, translation-ambiguous words may suffer from a decrease in associative strength, similar to the fan effect (e.g., Anderson, 1974). The fan effect assumes that as more facts are associated with a concept, each fact has decreased probability of occurrence and thus lower associative strength. As a result, the more facts that are connected to a concept, the longer individuals take to recall the concept. In the case of translation-ambiguous words, a learner would take longer to retrieve a translation if additional translations are linked to the same word.The majority of previous studies have examined the one-to-many direction of ambiguity, where a single L1 word was mapped onto two foreign language vocabulary items (for an exception focusing on many-to-one mapping, see

Novel Word Learning Among Bilinguals Can Be Better Through the (Dominant) First Language Than Through the Second Language

“…Therefore, it might be the case that when proficiency in the language through which learning takes place is higher, there are more cognitive resources available for processing of the newly acquired information (Segalowitz, 2003; Tzelgov & Kadosh, 2009), especially early in learning. This proficiency/resource‐availability proposal is supported by the general observed association between proficiency in the L1 and learners’ ultimate proficiency in the L2 (e.g., Cummins, 1979), and more directly by a recent study in which proficiency in the language through which learning took place modulated learning (Degani & Goldberg, 2019). In that study, Degani and Goldberg compared native Hebrew speakers to Russian–Hebrew multilinguals in learning Arabic words and did not find a multilingual advantage when learning through Hebrew (the L2 of the multilinguals).…”

“…Sample size was determined based on previous vocabulary learning studies in which item type and participants’ characteristics were observed (Degani & Goldberg, 2019; 30 participants per group, 12 items per condition). Nonetheless, we acknowledge that the estimated power to detect an interaction (of Cohen's d = 0.3) between LOI and item type with 29 participants per group and 15 items per condition is rather low (power = 0.531, using PANGEA; Westfall, 2016).…”

“…The learning phase in the first session consisted of two types of training cycles (following Degani & Goldberg, 2019): a L1/L2‐to‐L3 word repetition procedure and a L3 retrieval (recall) attempt. In the first type of learning cycle, a fixation cross appeared at the center of a screen for 750 milliseconds followed by a 250‐millisecond blank screen, which was then replaced by either the Hebrew (L1) or the English (L2) written translation (depending on the LOI condition), presented for 1,000 milliseconds, and followed by the auditory presentation of the German (L3) word via headphones.…”

“…(2019) suggested that for multilinguals who have experienced a switch in dominance such that they have become more dominant in their L2 rather than in their L1, a learning advantage may be observed when these multilinguals are learning through their dominant language, because they are experienced in regulating this dominant language in the course of using their first‐acquired, less dominant language. However, in the study by Degani and Goldberg (2019) all participants learned the Arabic foreign language vocabulary through what was in fact their dominant language (Hebrew), even though it was the L2 for those Russian–Hebrew speakers, and yet no multilingual advantage emerged. It is therefore possible that the order of acquisition of the languages does play an important role and that a learning advantage for bilinguals over monolinguals is limited to cases where the LOI is the first‐ acquired language irrespective of its dominance (Degani & Goldberg, 2019).…”

“…However, in the study by Degani and Goldberg (2019) all participants learned the Arabic foreign language vocabulary through what was in fact their dominant language (Hebrew), even though it was the L2 for those Russian–Hebrew speakers, and yet no multilingual advantage emerged. It is therefore possible that the order of acquisition of the languages does play an important role and that a learning advantage for bilinguals over monolinguals is limited to cases where the LOI is the first‐ acquired language irrespective of its dominance (Degani & Goldberg, 2019). To prevent this complexity from potentially confounding the current study, as the first study to test the effects of LOI on learning within the same population, we ensured that participants’ more dominant language was also their first‐acquired one.…”