International Open Access Journal Platform

1 Introduction

The global spread of English has inevitably led to its nativization and hybridization in diverse contexts, giving rise to World English. In China, one prominent manifestation of this contact is “Chinglish”1—a term often used pejoratively to describe English influenced by Chinese syntax, semantics, pragmatics, and cultural norms (Li, 1993). Although Chinglish is sometimes stigmatized as “broken English”, it is a common phenomenon that can be observed in learners’ output, signage, translations, and even in the speech of skilled bilinguals in certain contexts. This paper argues that it is not enough to view Chinglish from the perspective of error, which obscures the underlying cognitive processes that shape Chinglish. Instead, from a psycholinguistic perspective, we can understand Chinglish as a result of the basic ways in which bilingual minds process, store, and acquire linguistic information.

Psycholinguistics is the study of the psychological and neurobiological factors involved in the acquisition, use, and comprehension of language in humans, and it provides a powerful tool for analyzing second language (L2) production. Psycholinguistic research focuses on the mental representations and processes involved, such as lexical access, syntactic planning, and interactions between languages in the bilingual brain. By applying psycholinguistic theories (including language transfer, bilingual memory modeling, and sentence processing modeling under cognitive constraints), we can discover the systematic cognitive roots of Chinese-English features. Building on core psycholinguistic frameworks—language (especially conceptual) transfer, bilingual memory representation (e.g., the Revised Hierarchical Model), sentence processing strategies (e.g., the Competition Model), and the critical role of cognitive constraints in real-time production— this paper aims to elucidate how Chinglish emerges from the dynamic interaction between the deeply entrenched L1 (Mandarin) system and the developing L2 (English) system. Crucially, it argues that the pervasive influence of L1 patterns is profoundly amplified under the pressures of online communication, a phenomenon best explained through the lens of cognitive load theory.

2 Defining Chinglish: Systematic Cognitive Features vs. Transient Learner Errors

Chinglish encompasses a range of linguistic phenomena that stem from the influence of the Chinese language on the use of English. At the lexical level, it manifests itself in the direct translation of Chinese idioms or compound words (e.g., the direct translation of 丢脸 diū liǎn to “lose face”), semantic extensions based on Chinese concepts (e.g., using “open” to express 开灯 kāi dēng when it should actually be “turn on”)(Pavlenko, 1999), and creating new compound words (e.g., “air pork” for 航空猪肉hánɡ kōnɡ zhū ròu may refer to the pork served in airline meals).

At the syntactic level, common features include the omission of articles (e.g., “He is teacher”)(Kroll & Stewart, 1994), plural markers (e.g., “two book”), or prepositions where necessary, and the adoption of a non-standard word order that reflects the syntactic structure of Chinese (e.g., “This book I very much like” mimics the Chinese topic-comment structure 这本书我很喜欢 Zhè běn shū wǒ hěn xǐhuān); and overuse of certain common Chinese structures (e.g., redundant conjunctions like “although... but...” mirroring 虽然……但是…… suīrán... dànshì...).

At the pragmatic level, this is manifested by bringing Chinese directness into English and being perceived as blunt (e.g., “You must come tomorrow” instead of the more euphemistic “Could you come tomorrow?”), formulaic translation of Chinese polite phrases and organizing English discourse according to Chinese rhetorical patterns.

The key to understanding Chinglish lies in distinguishing (a) transient developmental errors arising from incomplete L2 knowledge from (b) systematic, cognitively entrenched features driven by L1 conceptual and processing patterns. The phenomenon of Chinglish, particularly its stable or systematic features, often arises not from a deficit but from the prioritization of communicative efficiency and, more fundamentally, the deep-seated cognitive mechanisms governing bilingual language processing. Understanding these psycholinguistic roots (explored in Section 3) is key to distinguishing transient errors from systematic L1-influenced patterns. Furthermore, recognizing this cognitive basis underpins the academic discourse on “China English” as a potential nativized variety, where select stabilized features may become conventionalized (Xu, 2010; He & Li, 2009).

3 The Psycholinguistic Basis for the Formation of Chinglish

The formation of Chinglish can be traced back to several core psycholinguistic principles at work in the minds of bilinguals:

3.1 Language Transfer and Conceptual Transfer: The Deep Imprint of L1

Many of the features of Chinglish are not accidental errors but are deeply rooted in a fundamental way our brains work when dealing with bilingualism: the deep imprint left by our native language (L1) continues to influence the learning and use of our second language (L2). This phenomenon is called transfer by linguists (Odlin, 2005). Simply put, transfer means that what you already know about a language (especially your native language) is “transferred” to the learning and expression of the new language. But the key understanding is that the effect is much more than simply replacing Chinese words with English words that sound like them. Psycholinguistic research reveals that deeper and more persistent is “conceptual transfer”, that is, cognitive schemata that we have constructed through our mother tongue unconsciously shape the way we understand and express the world in our second language.

This conceptual level of transfer is everywhere in Chinglish and profoundly affects lexical choices and grammatical constructions. A typical example is the use of verbs. Chinese verbs often have broader semantic ranges and different combinatorial restrictions than their English “equivalents”(Jarvis, 2007). For example, the Chinese 开 kāi is a powerful «multi-faceted» verb that can be used to open a door or window (physically open it), to start an electrical appliance (e.g., turn on a light, turn on a TV), or even to start a meeting or start a car. In English, however, more refined «tools» are needed: open for opening a door, turn on/switch on for turning on a light, start/hold for holding a meeting, and start/drive for driving a car. When learners map the broad concept of «start function» in Chinese directly onto the English verb open, which they first learned or are most familiar with, the expression “open the light” arises. Similarly, “close the TV” from the Chinese 关 guān applies to both physical closure (shutting the door) and functional closure (turning off an appliance). This is not essentially a case of choosing the wrong word but of the brain constructing meaning while its deeper conceptual blueprints are still in Chinese.

Conceptual transfer also creates barriers to the expression of time. The expression of time in Chinese mainly relies on vocabulary (e.g., “yesterday”, “being”) and aspect markers indicating the state of action [e.g., “了” for perfective aspect (indicating completed action), “着” for continuity]. It does not have an obligatory and systematic grammatical framework of “tense” like English, i.e., the verb changes-ed, -s or the auxiliary verb will/have must be used to indicate past, present, future, or completion. This fundamental conceptual difference means that for native speakers of Chinese, “time” is understood primarily through contextual vocabulary and action states rather than being forcefully encoded in the morphology of the verb itself. Therefore, when speaking English, even if they know the rules, they often ignore the tense markers and say sentences like “Yesterday I went to school”. It is even more difficult to master the complex English perfect tenses (e.g., have done) because the Chinese language lacks a fully corresponding, grammatical conceptual framework for “completion”.

Another significant difference is in how things are viewed in terms of “countability” and “definiteness”. Chinese nouns themselves do not usually have an obligatory singular-plural distinction (one book/two books, the “book” itself remains the same), and they do not need articles such as English a/an/the to mark whether they are generic or specific. This means that the concepts of “a book” and “that book” are clear and natural in Chinese, without the need for additional grammatical markers. When this conceptual convention is transferred to English, learners tend to ignore the grammatical signal of “quantity/certainty” required by English, resulting in the omission of articles (e.g., “He is a teacher”) or incorrect pluralization (e.g., saying “information” as “information”). These “errors” reflect the fact that the brain has not yet fully internalized the mandatory grammatical “counting” and “definiteness-marking” modes of thinking in English when dealing with noun concepts.

The expression of spatial relationships is also affected by conceptual transfer— Chinese and English use different “linguistic toolkits” when describing the position or direction of objects. For example, Chinese sometimes prefer to associate directional or path information directly with verbs (e.g., “进” means to enter, and “上” means to go up). In contrast, English relies more on prepositions ( i.e., in, on, at, about) to precisely label the relationship. When learners apply the collocational requirements of Chinese verbs directly to English verbs, prepositions such as “discuss the problem” and “emphasize this point” will be used. When the collocation requirements of Chinese verbs are directly applied to English verbs, prepositions like “discuss the problem” or “emphasize this point” will appear. The reason behind this is that native language conventions still dominate the conceptual path of dealing with spatial and abstract relations.

At the heart of why these transfers are so pervasive and persistent is the fact that the conceptual categories and linguistic processes we have built up through our mother tongue are highly automated linguistic thinking that has been reinforced through decades of daily use. They have become conceptualizing and expressing the world. When we learn or use a second language, especially in situations that require quick responses, expression of complex ideas, or stress, the brain instinctively prioritizes the most familiar and fluent native language to convey core concepts. It is not only slow but also error-prone to try to forcefully switch to a second language, min, and that is not yet proficient and requires more cognitive resources to construct. As a result, even when learners know the correct form of the English language, that deeply ingrained native conceptual path still often prevails in real language output, giving rise to a wide variety of Chinese English expressions. It’s important to understand that many features of Chinglish stemming from conceptual transfer represent a natural cognitive strategy for efficiency. This reliance on entrenched L1 pathways for efficiency extends beyond conceptual mapping to the core mechanisms of real-time language processing: lexical access (Section 3.2) and syntactic assembly (Section 3.3). The following sections explore these critical processing mechanisms: the lexical-conceptual pathways governing vocabulary choice (Section 3.2, RHM) and the cue-weighting strategies driving syntactic production (Section 3.3, Competition Model).

3.2 Bilingual Memory Representation: The Revised Hierarchical Model (RHM)

Kroll and Stewart’s Revised Hierarchical Model (RHM) provides a central insight into why Chinese learners often show traces of “Chinglish” in their English vocabulary choices and collocations. The model reveals a key fact: there is a profound asymmetry in the learner’s brain between the native language (L1) vocabulary, the second language (L2) vocabulary, and the concepts they represent, rather than an equal connection (Kroll & Stewart, 1994). For the vast majority of beginners and even intermediate learners, this is often the case: an English word (L2) first strongly activates its corresponding Chinese translation (L1), which then activates the concept it represents (Kroll & Stewart, 1994). In contrast, the connection between the Chinese word itself and the concept is direct and much stronger. It is only as English proficiency continues to increase that English words can gradually build up a “shortcut” to concepts, freeing them from their dependence on Chinese translations.

This mediated approach, which relies on Chinese words as a bridge to understanding English words, is the breeding ground for many typical Chinglish vocabulary problems. When you need to use an English word (e.g., “light”) to express something, your brain will probably activate its Chinese equivalent, “灯dēng”, first (Kroll & Stewart, 1994). The trouble is, the word “light” is not alone in Chinese; it comes with a set of customary partners and usages. For example, we often say “开灯kāi dēng”. So, when searching for common collocational verbs for “light”, “开kāi” is strongly activated. As a result, the brain logically outputs this Chinese collocation habit together to form the Chinglish expression “open the light”. The key point is that this error is not due to not knowing the phrase “turn on”, but rather, under the pressure of real-time and rapid language output, the most familiar and automated pathway (L2->L1->concept/collocation) is prioritized, while the more English-compliant phrase “turn on”, which requires extra cognitive effort to retrieve, is activated. “turn on” was temporarily bypassed.

When it comes to more difficult situations, i.e., concepts, idioms, or expressions that are specific to the Chinese language do not have a perfect counterpart in English (lexical gaps), the path dependency revealed by RHM can lead to another type of typical Chinglish: hard or direct translation (Jiang, 2000). For example, when trying to express the encouragement to cheer up “加油jiāyóu”, the learner knows the core concept (encouragement and support) but cannot find a concise and relevant equivalent in the English lexicon (e.g., “Come on!”, “Go for it!”, “You can do it!”). At this point, the brain’s quickest and most direct response is to follow the ingrained path: activate the Chinese word “加油 jiāyóu” and literally break it down into “add” and “oil”, then combine them with known English words, thus “Add oil” is born.

In sum, RHM clearly depicts a core cognitive pathway for the production of lexical-level Chinglish: a strongly mediated connection from L2 to L1 to concept/collocation. This explains why, even after learning correct English expressions, formations that have been influenced by the native language often persist in real, pressurized exchanges, because they are rooted in the brain’s most primitive and automated ways of connecting bilingually. Overcoming these problems requires not only building vocabulary but also conscious, extensive practice in how to root English words in the concepts themselves and familiarize them with the way they are naturally combined in English contexts, ultimately reducing reliance on the mediating L1 translation route.

3.3 Sentence Processing: The Competition Model

Building on the role of L1 conceptual structures (Section 3.1) and lexical access pathways (Section 3.2), the Competition Model proposed by Bates and MacWhinney provides a powerful framework for understanding how sentence-level grammatical features in Chinglish arise from the transfer of L1-based processing strategies (Bates & MacWhinney, 1989). Specifically, it explains why learners develop specific syntactic patterns by examining how they assign weights to different cues (like word order vs. morphology) when constructing sentences under processing constraints. The core idea of the model is that the brain relies on and weighs a number of different cues when we understand or construct a sentence and need to determine “who did what to whom” (i.e., assigning grammatical roles such as subject, object, etc.). These cues include the order in which words appear (word order), variations in word morphology (e.g., subject-verb agreement, pronoun agreement), and the plausibility of the meaning of the words themselves (semantic plausibility). The key lies in the fact that different language communities give different “credibility weights” or priorities to these clues in their long-term use.

When dealing with sentences, Chinese attaches great importance to two major clues: the fixed order of words (especially the strict “subject-verb-object” SVO structure) and the natural collocation of word meanings (semantic rationality). For example, in the sentence “The cat chased the mouse”, the cat comes first, and it is reasonable that the action of “chasing” is given by the cat, so it must be the subject; “the mouse” comes second and is suitable for the subject. Chinese seldom rely on fine morphological markers like verb conjugation (e.g., he walks* in English) or pronoun lattice (he vs. him) to label roles because these cues are inherently underdeveloped or non-existent in Chinese.

The problem is that when Chinese learners turn to English, the language system in their brains, which their mother tongue has profoundly shaped, does not automatically reset. They tend to use the most efficient and reliable set of judgment criteria in Chinese to process English sentences. This leads to several typical phenomena in Chinglish:

Morphological markers are often ignored (Bates & MacWhinney, 1982): Since Chinese hardly relies on verb endings (-s, -ed) or pronouns (he/him) to determine who is the subject and who is the object, it is natural for learners to be insensitive to, or under-appreciate, these crucial morphological cues in English. As a result, errors like “He went to school yesterday” (missing the third-person singular -s and the past-tense -ed) or confusing “He saw she” (using the subject pronoun “she” instead of the object pronoun “her”) are common. In high-pressure, real-time language processing, the brain instinctively prioritizes what it considers to be the most central and reliable cues (word order and semantics in Chinese) and treats those morphological variations that are irrelevant in the native language as secondary, temporarily sacrificed “decorations”.

Word order becomes unusually rigid: Chinese’s strict reliance on SVO order as a central cue makes learners very uncomfortable with the permissible, even common, variations in word order in English. For example, the English passive voice (e.g., The book was read by him) breaks the SVO order of the main verb (e.g., He read the book). As well as relational clauses or certain gerund positions, the basic order is also adjusted (e.g., The man whom I saw yesterday...). Learners either try to avoid these structures or produce them with unnatural word order or phrasing. The root cause is that changing the familiar SVO word order is like removing the most important signposts in their processing system based on the weighting of cues in their native language, which greatly increases the difficulty of comprehension and the cognitive load of the production and, thus, tends to favor the safest SVO direct form.

Preference for the “topic-first” structure: Chinese has a very natural habit of placing the object to be talked about (the topic) at the beginning of the sentence and then describing it (the comment), e.g., “I like this book very much.” The reason why this “topic-comment” structure is so efficient in Chinese is that it perfectly fits the Chinese processing mechanism that relies on the two core cues of order (the topic comes first) and semantics (the comment part describes the topic). When learners process English, this efficient pattern is easily transferred to produce sentences like “This book, I really like it”. Although English has a similar structure (Topicalization), its frequency of use, context and acceptance are different from Chinese. Learners choose this structure for the deeper reason that it directly invokes and conforms to the most ingrained and efficiently processed native sentence construction blueprint in their brains.

In conclusion, the competing models reveal that many of the syntactic features of Chinglish are not simply “forgetting the rules” but are natural extensions of the highly automated sentence processing strategies (cue weighting) that learners have built up in their brains based on their native language experience in cross-language communication in the English context. Overcoming these features requires not only learning the English rules themselves but also consciously adjusting the brain’s prioritization and trust in cues when processing sentences.

3.4 Cognitive Load and Processing Constraints

The psycholinguistic mechanisms explored in Sections 3.1 to 3.3— the pervasive influence of L1 conceptual structures (Conceptual Transfer), the asymmetrical lexical access pathways favoring L1 mediation (Revised Hierarchical Model), and the entrenched L1-based cue-weighting strategies for sentence processing (Competition Model) – provide the fundamental cognitive blueprints for the formation of Chinglish features. However, the extent to which these L1-influenced pathways dominate actual language output, particularly in spontaneous communication, is critically moderated by the constraints of cognitive resources and processing demands during real-time production. In essence, cognitive load theory serves as the crucial explanatory framework for when and why the L1-based mechanisms described earlier prevail over target-like L2 forms, even when learners possess declarative knowledge of the latter.

Language production itself is an extremely cognitive, resource-intensive process. Imagine a real conversation or writing in which the brain has to perform multiple demanding tasks simultaneously under time pressure: conceiving the core meaning to be expressed (planning meaning), searching for appropriate words from memory (retrieving vocabulary), assembling these words into sentences according to rules (constructing syntactic structures), and at the same time checking that what is said or written is grammatical and contextually correct (monitoring the output). As Just and Carpenter (1992) point out, we have a finite working memory capacity, like a processor with an upper limit on its processing power, which can easily become overloaded when the task is too complex or time is of the essence.

When cognitive resources are pushed to their limits - for example, when a learner is trying to make a complex point, is in a tense conversational situation, or needs to multitask - the brain instinctively activates “survival mode”: prioritizing core meaning conveyance at the expense of formal accuracy (Kroll, Michael, Tokowicz, & Dufour, 2002). Why does this “meaning over formation” trade-off occur? Because conveying the core intention is the fundamental purpose of communication, while formally accurate modifications often require additional cognitive effort. At this point, language processing patterns and structures that have become highly automated through years of use of the native language (L1), with little mental effort, become the most convenient and resource-efficient tools. In contrast, the conscious invocation and construction of entirely new expressions that conform to the norms of a second language (L2) consumes a great deal of valuable cognitive resources. This leads to a key phenomenon: even if learners know what the correct way to say a certain sentence in English is, in high-pressure situations, their brains still prioritize the less laborious “old way” of mapping native language structures directly onto English. This is one of the core cognitive motivations behind many of the Chinglish constructions that appear to be “knowingly” produced.

More problematically, such resource constraints can severely weaken the brain’s “quality control department”—output monitoring. Effective monitoring, i.e., detecting and correcting errors in expression (especially those in irregular forms influenced by the native language), inherently requires spare attentional resources for fine-grained comparison and adjustment. However, when the cognitive system is already operating at full capacity to support basic expressive tasks, the extra attention required for such “quality control” is stretched to the limit or even squeezed out altogether. Consequently, effective output monitoring is impaired, allowing L1-influenced forms to emerge unchecked.

To further “lighten the load”, learners often rely on a cognitive “crutch”— the overuse of pre-memorized chunks of language or fixed phrases translated directly from their mother tongue. Whether it’s memorized conversation templates (e.g., the stereotypical response to the greeting “Fine, thank you, and you?”) or word-for-word translations of Chinese expressions into English (e.g., translating “好好学习天天向上hǎo hāo xué xí tiān tiān xiànɡ shànɡ” to “Good good study, day day up”), these formulaic linguistic units are essentially a cognitively economical strategy. Because they do not require complex lexical search and syntactic construction in the present, they can simply be retrieved as a whole, greatly reducing the burden of real-time communication processing. Unfortunately, these directly transplanted phrases often ignore the English language habits and cultural contexts, which, although relieving the cognitive pressure at the moment, become a typical expression of Chinglish at the pragmatic level and may solidify unorthodox expression habits.

When the brain fails to retrieve even the right words under high pressure, the simplification strategy becomes the last line of defense: the brain activates high-frequency words with broad semantics (e.g., “open” for all opening actions), discards non-essential morphological markers (omitting articles, simplifying tense marking), or retreats to simple sentences closest to the native pattern. These seemingly lazy choices are, in fact, rational compromises in the face of cognitive overload - trading expressive coarseness for information transfer possibilities. The latest research provides hard evidence of this: Chinese learners’ working memory capacity is significantly and positively correlated with English grammatical accuracy in time-limited writing. This is direct evidence that the more scarce the cognitive resources, the more pervasive the Chinglish features (Guo & Peng, 2021).

4 Implications and Moving Forward

An in-depth understanding of the above psycholinguistic roots has far-reaching implications for teaching practices and language attitudes. A fundamental shift is needed at the teaching level: from error correction to revealing conceptual differences (e.g., the semantic range of Chinese and English verbs, spatio-temporal conceptualization), and deepening cognition through comparative analysis; establishing a direct link from L2 vocabulary to concepts and weakening the L1 mediation by means of immersive input (extensive reading/listening) and promoting L2 conceptual mediation through activities like contextualized interpretation and meaning-based processing in English (interpreting meanings in English and contextualizing)(Jiang, 2000); cultivating metacognitive awareness, helping students to understand the causes of Chinese-style English (e.g., migration mechanisms, cue weighting), and develop autonomous monitoring strategies; scientifically manage cognitive load, design scaffolded tasks to enhance proceduralization/automatization of L2 knowledge, allow planning time before complex tasks, and create low-pressure output environments; and explicitly teach the rules of English pragmatics to avoid the direct transplantation of Chinese politeness strategies.

At the socio-cognitive level, there is a need to recognize that Chinglish is a natural product of bilingual cognition rather than a deficiency in proficiency. This will help to build a more inclusive attitude, shifting the focus from “removing traces of Chinese” to “guiding language development based on an understanding of cognitive mechanisms”, pursuing higher levels of proficiency and rationalizing the stable features that may be embedded in “China English”. In the future, we can combine eye-tracking and electroencephalography (EEG) with Event-Related Potentials (ERP) analysis. Other methods to accurately measure the processing of specific Chinglish features, and with the help of corpus analysis, we can track the evolutionary paths of these features in the development of bilingual speakers’ language proficiency and the potential trend of their normalization.

5 Conclusion

Chinglish is much more than a collection of random errors. The formation of Chinglish is rooted in the complex interplay of bilingual cognitive processes, which can be revealed through psycholinguistic research. The pervasive influence of entrenched L1 conceptual structures and processing procedures, as described by conceptual transfer theories, radically interacts with the structure of bilingual memory representations as described by models such as RHM. The Competition Model explains how syntactic choices reflect the transfer of cue-weighing from Chinese to English sentence processing. Most importantly, these processes take place within the constraints of limited cognitive resources. Hence, reliance on automated L1 patterns is a natural and often unavoidable strategy during real-time communication, especially at lower levels of proficiency.

This psycholinguistic perspective on Chinglish unravels the mystery of Chinglish and challenges the purely deficit-oriented view. It emphasizes the active cognitive engagement of language users, navigating the complex terrain between the two language systems. While the goal of Chinese English pedagogy is usually to develop competence in line with international standards, recognizing the psycholinguistic realities of the formation of Chinglish is crucial to the development of effective, cognitively informed teaching strategies. It encourages a focus on establishing robust second language conceptual representations and processing procedures, fostering metalinguistic awareness, and managing cognitive load in second language production. In the final analysis, understanding the reasons behind Chinglish can pave the way for a more nuanced understanding of Chinese-English bilingualism and the continuous evolution of English in the Chinese context.

References

[1] Bates, E., & MacWhinney, B. (1982). Functionalist approaches to grammar. In E. Wanner & L. R. Gleitman (Eds.), Language acquisition: The state of the art (pp. 173-218). Cambridge University Press.

[2] Bates, E., & MacWhinney, B. (1989). Functionalism and the competition model. In B. MacWhinney & E. Bates (Eds.), The crosslinguistic study of sentence processing (pp. 3-73). Cambridge University Press.

[3] Guo, T., & Peng, D. (2021). Working memory capacity and article use in Chinese EFL learners’ online production. Language Teaching Research, 25(5), 712-731.

[4] He, D., & Li, D. C. S. (2009). Language attitudes and linguistic features in the “China English” debate. World Englishes, 28(1), 70-89.

[5] Jarvis, S. (2007). Theoretical and methodological issues in the investigation of conceptual transfer. Vigo International Journal of Applied Linguistics (VIAL), 4, 43-71.

[6] Jiang, N. (2000). Lexical representation and development in a second language. Applied Linguistics, 21(1), 47-77.

[7] Just, M. A., & Carpenter, P. A. (1992). A capacity theory of comprehension: Individual differences in working memory. Psychological Review, 99(1), 122-149.

[8] Kroll, J. F., & Stewart, E. (1994). Category interference in translation and picture naming: Evidence for asymmetric connections between bilingual memory representations. Journal of Memory and Language, 33(2), 149-174.

[9] Kroll, J. F., Michael, E., Tokowicz, N., & Dufour, R. (2002). The development of lexical fluency in a second language. Second language research, 18(2), 137-171.

[10] Li, W. (1993). China English and Chinglish: Reflections on the emerging English co-existing in China. Foreign Language Teaching and Research, (4), 18-24.

[11] Odlin, T. (2005). Crosslinguistic influence and conceptual transfer: What are the concepts? Annual Review of Applied Linguistics, 25, 3-25.

[12] Pavlenko, A. (1999). New approaches to concepts in bilingual memory. Bilingualism: Language and Cognition, 2(3), 209-230.

[13] Xu, Z. (2010). Chinese English: Features and implications. Hong Kong University Press.