Code switching & Code meshing

Figure 1. Language proficiency and lexical access

          One particular translanguaging phenomenon of multilingual speakers is code switching and code meshing. It is believed that multilingual speakers have all of their language codes integrated in one repertoire, and when communicating, all the linguistic codes are activated together. Thus, code mixing/translanguaging is considered as naturally inevitable practices. On the other hand, since both languages are potentially active and competing to control output, Applied linguists of bilingualism indicated that language switching takes time because (1) it involves a change in language schema for a given task (schema level inhibition), (2) any change of language involves overcoming the inhibition of the active lemmas with non-target tags (tag inhibition) (Green, 1998, p. 73); (3) cost of switching is asymmetric, and switching into dominant codes costs more than the reverse direction; and (4) the bility to engage in fluent code switching is a hallmark of high proficiency in two languages (Miccio et al. 2009), given that successful and fluent code switching requires a high degree of knowledge of and sensitivity to the grammatical constraints of both languages (kroll et al, 2015).
          If our brain is prone to take the most economic course when
communicating, and if code switching and code meshing cost cognitive capacity, they may not always be the economic selection of multilingual speakers. I hypothesize that some codes may involve more switch cost (e.g. language specific codes, numeral codes, & functional lexical lemmas)  but some may be triggered more spontaneously (e.g. shared content lexical lemmas) (See Figure 2); different language proficiency may result in different cost of language switch (See figure 1); explicit training of code switch may reduce switch cost; contexts that allow more flexible code mixing can reduce unpredictable switching and lead to less lexical suppression (for the more suppressed codes take longer to switch into); and code-switching/meshing from L2 to L1 when composing L2 writing can be ineffective, especially for high proficiency students who have direct access to L2 codes but need to intentionally use L1 during L2 production.

The Modified Hierarchical Model. A. Pavlenko (2009), Conceptual representation in the bilingual lexicon and second language vocabulary learning. In A. Pavlenko (ed.), The bilingual mental lexicon: Interdisciplinary approaches , p. 147. Multilingual Matters, Buffalo, NY. 

          I also would like to propose a broader definition of code-meshing that it refers to meshing two or more codes as a whole at not only the level of lexicon, but phonology, morphology, syntax and semantics. For example, Chinglish, under this definition, can be seen as a kind of code-mashing of Chinese and English, which has Chinese syntax and semantics as the base or matrix and bring in English lexemes as the "guest" (e.g.  " I loved a girl at my first sight" ). Given this definition, all the multilingual speakers' language productions are translingual and code-mashed.

Teaching Implications:
1. Since L2 may affect L1, EFL students when writing in L2 dominant contexts may experience reduced access to their L1. I have students complained that code-switching during L2 writing sometimes was not effective. However, most students reported that using L1 to plan before writing is effective. This finding suggests that before the target L2 task schema is activated, L1 can be of great help in idea generation, logic reasoning, and detail elaborations. However, once the L2 task schema is activated, and L1 is inhibited during L2 writing, L1 code-switching may be less effective or even lead to more errors. Therefore, (1) Pre-writing contexts (L1 dominant): teachers can create an L1 contexts triggering L1 writing schema to help students take good use of their L1 resources. (2) Writing contexts (L1 and L2): Do not require students to use certain language when writing. Instead, allowing students to use whatever languages during their processes of drafting. This can allow students to spontaneously use the more strongly activated codes through competing. (3) Post-writing (L2 only): When making final revision/editing, students can then encouraged to use the target language only.
2. EFL writers should not only learn to regulate and control cross-language competition but also learn to shuttle between all the language resources in their repertoires in order to creatively negotiate or mash languages for writing agendas.
3. EFL students with different English proficiency levels can be benefited from L1 differently. Low proficiency writers may be benefited from L1 in lexical search, idea generation, logical reasoning, translating, and monitoring. However, high proficiency writers may prefer to use L1 in idea generation and monitoring. Other usages of L1 during L2 writing may lead to interference for high proficiency writers unless L1 can compensate their situational writing needs.