The Revised Hierarchical Model: Explicit and Implicit Memory

Abstract

We first introduce three influential bilingual memory models that include the word association model, the concept mediation model, and the Revised Hierarchical Model (RHM).

Appendices

List of Keywords and Concepts

Automatic activation, Bilingual memory models, Concept mediation model, Conceptual explicit memory test, Conceptual representation, Conceptual processing, Conceptual tasks, Exemplar generation, Explicit memory, Explicit memory contamination, False memory effect, Hierarchical models, Implicit memory, Lexical representation, Mask manipulation, Prospective expectancy generation, Revised Hierarchical Model (RHM), Translation asymmetry, Translation priming, Word association model, Word association model

Thought Questions

• 1. How would the manipulations that encourage explicit contamination (e.g., a production task with a high proportion of test items that had appeared in the study phase) influence bilinguals’ long-term translation priming effects in the L1-to-L2 and L2-to-L1 directions? Would this depend on their L1 and L2 proficiencies?

• 2. How would the manipulation that encourages prospective expectancy generation (e.g., long SOA and prime duration) influence bilinguals’ short-term translation priming effects in the L1-to-L2 and L2-to-L1 directions? Would this depend on their L1 and L2 proficiencies?

• 3. This chapter demonstrates the possibility of applying the RHM to explain the encoding and retrieval processes involved in bilingual implicit and explicit memory. How would the explanation provided by the RHM be different from those provided by traditional memory theories (e.g., transfer appropriate processing)?

• 4. How would students’ knowledge acquisition be influenced by encoding/study and retrieval/test languages in real-life classroom settings? Besides basic vocabulary acquisition, how would the effect of language-match influence students’ higher-order cognition, such as analogical reasoning and critical thinking?

• 5. Following up Marian and her colleagues’ works on language-dependent autobiographical memory, would this effect depend on bilinguals’ L2 proficiency at the time when the memories are encoded (i.e., in the past) or when the memories are retrieved (i.e., at present)?

• 6. Would the expectancy of test type modulate bilinguals’ strategies in encoding the study items (e.g., emphasizing the form/language or the meaning) and in turn, the effect of language match on their memory performance?

Applied Issues in Learning and Memory in the Acquisition of an L2 and Vocabulary Learning

Previous bilingual implicit memory studies suggest that the long-term translation priming effect more likely occurs when the task taps conceptual processing (e.g., animacy decision and exemplar generation) than when the task taps only lexical processing (e.g., lexical decision). That is to say bilinguals could recognize the translation equivalent of a word more easily following a presentation lag of that word when they activate the meaning of the words, rather than judge word lexicality based on familiarity (e.g., lexical decision, see Balota and Chumbley 1984) . This highlights the importance of activating word meaning, instead of memorizing the lexical forms in a rote fashion (e.g., spelling or pronunciation), when we learn new vocabulary in a second language. This suggestion supports one of the recommendations made by researchers in second language vocabulary acquisition (see Nation 2001, for a review).

Previous bilingual explicit memory studies suggest that bilingual memory would be better when the words are encoded and then retrieved in the same language than when they are encoded and then retrieved in two different languages. It is important to investigate whether this relationship could generalize to real-life classroom settings. For example, would bilingual students remember better in their L1 than in their L2, regardless of their L2 proficiency? Would bilingual students’ memory performance be lower when the concepts are taught in one language (e.g., their L1, Chinese) but are then subsequently tested in another language (e.g., their L2, English)? If so, would this happen only for bilinguals with low L2 proficiency, but not for those with high L2 proficiency? Clearly, the answers to these questions would have strong implications for policies on medium of instruction in educational settings. For example, should students with various L2 proficiencies be taught in L2 (see Tollefson and Tsui 2004, for a review) ? Hence, the findings reviewed in this chapter, despite having been based on laboratory tasks, could still have implications for applied settings.

Suggested Projects

• 1. Students could orthogonally manipulate the nature of the test type (implicit vs. explicit), the level of processing at encoding (lexical/perceptual, e.g., vowel counting, vs. conceptual, e.g., pleasantness rating), the level of processing at retrieval (lexical/perceptual vs. conceptual), encoding/study language (L1 vs. L2), and retrieval/test language (L1 vs. L2) using the same set of experimental stimuli. The implicit and explicit memory tasks should conform to the retrieval intentionality criterion (Schacter et al. 1989; Tse and Pu 2012) ; that is, all stimuli and procedures in both tasks must be the same except participants’ intention to retrieve during the task, for example, explicit/implicit stem completion and explicit/implicit category exemplar production. This large-scale study would tease apart and examine the effects of test task and level of processing at encoding and retrieval on bilingual implicit and explicit memory, which has not been completely addressed in the literature.

• 2. Students could orthogonally manipulate the factors that encourage prospective expectancy generation (e.g., masked manipulation, SOA, and prime duration) and the type of tasks that tap different levels of processing (e.g., lexical decision vs. semantic categorization) to explore their influence on the short-term translation priming effect in two language directions (L1-to-L2 vs. L2-to-L1). The influence of bilinguals’ L2 (and even L1) proficiency on this interaction should also be taken into consideration.

• 3. By adapting the explicit and implicit DRM paradigm (e.g., Meade et al. 2007; Tse and Neely 2005, 2007) , the DRM lists could be used in the translation priming paradigm to explore the relationship between test type and encoding and retrieval languages. For example, immediately following the presentation of a list of DRM items, bilinguals would be presented and required to respond to a critical lure that can be in the same or different language as the list items. The decision could be affected by lexical/perceptual processing (e.g., lexical decision) or conceptual processing (e.g., animacy decision). This would inform us about the interaction between the language-match and the nature of tasks that tap more shallow vs. deep processing.

Appendix

Table 1 Selective review on the magnitude of L1-to-L2 and L2-to-L1 short-term translation priming effects using the lexical decision task with noncognate translations

Table 2 Selective review of memory performance in the between- and within-language conditions as a function of test type, instructions, and involvement of conceptual processing

Table 3 Selective review of the performance of the DRM list in false recognition memory tests