“How the Furby Coming is...”: Interference of First Language and Culture in Thai EFL Learners’ Paragraph Writing

This study investigated features of first language and cultural interference in Thai EFL learners’ English paragraph writing on popular culture. Drawing from theoretical grounds of interlanguage, language interference, and rhetorical interference, the sample of 30 English paragraphs of Thai EFL undergraduate learners was examined quantitatively and qualitatively. The English writing included 15 paragraphs from the Thai learners with high exposure to English language (TEH) group, and 15 paragraphs from those with the low exposure to English language (TEL) group. Using analysis models of metadiscourse markers and topical progressions, the findings revealed the preference of both groups in the use of interactive and interactional devices as well as SP, PP, and EPP types of topical progressions. The preference highlights the feature of oral-based, inductive, or reader-responsible writing orientation with a possibility of writing development, especially among the TEHs to reach expectation of the target language readers. The findings encourage assessing the Thai EFL learners’ writing as a process and raising frequent awareness of both language and rhetorical interferences when writing English texts. As the introductory stage during COVID-19 remote learning, writing to express learners’ interests could be used as an effective communication strategy for a positive instructor-learner relationship which assists the learners to further engage in the class in a more meaningful way.


Introduction
Many studies on L2 or English language learners' productions in both speaking and writing (e.g., Jaiprasong & Pongpairoj, 2020;Kazazoğlu, 2020;Kerz & Weichmann, 2015;Saito, Trofimovich, & Issacs, 2015) have revealed interference of L1 or learners' native languages and cultures. With the more degrees of differences between L1 languages and cultures and those of L2, the higher degrees of interference exhibit in the productions. This hindrance in L2 productions poses more difficulties for L2 learners to speak or write as target language audiences expect. In L2 writing, which is highly complex, dynamic, multicomponential, and multidimensional (e.g., Bulté & Housen, 2015), interference of L1 language could manifest itself in either positive or negative transfers. Positive transfers of the learners' L1 language facilitate L2 acquisition and learning while that of negative transfers impacts the learners' mastery of their L2 ( Bardovi-Harlig & Sprouse, 2017). The focus of previous studies on L1 language interference (e.g., Kazazoğlu, 2020;Watcharapunyawong & Usaha, 2013) has weighed more on negative transfers as English language learning challenges. Apart from the notion of L1 language interference, several scholars (e.g., Dita, 2009;İnceçay, 2015;Simpson, 2000) have investigated a great deal on L2 learners' cultural interference in formal writing tasks, such as abstract, explanatory, persuasive, or academic paragraph writings, entailing strict genres and expected English patterns as the learners' writing products. Some scholars (e.g., Carreon, 2006) have explored similar issue in informal writing tasks, such as free writing and personal journals which the findings suggest an opportunity for L2 learners to produce alternative authentic texts as the learners' writing processes reflecting their worldviews and language experiences.
In the Thai university context, Thai learners of English or Thai EFL learners usually write in a paragraph or an essay. Many research works in the mentioned context (e.g., Kongkaew & Cedar, 2018;Phoocharoensil et al., 2016) tend to capitalize the L2 audiences' expected forms or syntactic means of writing as to assess the Thai EFL learners' writing quality of formal writing tasks, especially academic paragraph writing. These studies appear to overlook the notions of rhetoric or "a method of organizing syntactic units into larger patterns" (Kaplan, 1966, p. 5) in the learners' English writing and of preferences of L1 culture-specific rhetorical patterns among the learners (Kaplan, 1972(Kaplan, , 1987. In addition, during COVID-19 remote learning and especially in the first few weeks of any English lecture courses in the Thai universities, Thai EFL learners are often in their passive or silent mode in which it may discourage effective communication between instructors and learners. Research studies related to pedagogical content knowledge have shown that teaching-learning context could directly impact both young and adult learners' levels of media literacy, critical thinking and language skills as well as expanding textual awareness when using popular culture or pop culture as a part of classroom materials (e.g., Tuzel & Hobbs, 2017). Forms of pop culture could include films, Netflix series, online games, music, clothing, food and drinks, sports accessories and many more which usually transmit through mass and social media. If the writing topic is assigned as to relate to pop culture, the Thai EFL learners may communicate through English paragraph writing with their instructors while constructing their pop culture knowledge together with their choices of words, clause structures, and overall writing organization. In the Thai university context, the parts in examining culture-specific rhetorical patterns or discourse-oriented English paragraph writing especially in informal writing tasks tend to receive less attention and thus remain to be explored. Many contrastive rhetoric scholars (e.g., Almuhailib, 2019;Colombo, 2012;Kaplan, 1966;Leki, 1997) have addressed that learners' L2 or English writings are culturally based. The L2 learners' culture-specific rhetorical patterns of writing could be judged as inconsiderate with the patterns expected by the target culture's readers due to the distinction between high-context and low-context cultures Hall, 1976). On the one hand, writers in a high-context culture, such as in Chinese, Japanese, and Thai cultures, usually construct their writing in loose organizational patterns that the shared meaning in the text is presumed (Rubin, Goodrum, & Hall, 1990). With the view of cooperative and active readers in mind, this high-context culture could also be recognized as an oral-based culture (e.g., Ong, 1980Ong, /2006 in which the readers are perceived as to share a common view of the world with the writers, to be active seekers of meaning, and to derive at mutual understanding of the text. On the other hand, authors in a low-context culture, such as in North American cultures, shape their writing to accommodate the readers with close organizational patterns that presuppose the author's effort of clarity and considerateness. With this view of writer-responsible texts, the low-context culture is known as a literary-based culture in which the writers are perceived as the actor or the leader of the text. Through the views of contrastive rhetoric, L2 learners' discourse organizational patterns in writing could be categorized into two orientations. They are literacy-based and oral-based orientations. Differences in features of oral-based rhetorical patterns of writing and those of literary-based ones could be described in the following continua as displayed in table 1. Table 1. Comparison and contrast between features of oral-based and literary-based rhetorical patterns of writing based on Kaplan (1966), Purves and Purves (1986) and Rubin, Goodrum, and Hall (1990) cohesive ties other than repetition, substitution, or collocation. On another end, a prepositional structured text admits little opportunity for the addition or deletion of material. Narrating or Dramatizing-Characterizing continuum refers to distinctions between the low-context writers whose emphasis are on characters and action; and the high-context ones whose degree of emphasis are on narration and dialogue-a distinguish feature of story-telling style. Message-focused and Reader-focused continuum refers to a degree in specifying a clear audience or the writer acknowledges the existence of that audience in which the latter has a clearer focus of audience.
With the distinction of the two kinds of cultures affecting discourse organizational patterns in writing, culture-specific rhetorical patterns of writing, especially among the high-context L2 writers have been proven to interfere with comprehension of readers in the low-context culture. However, many research studies have echoed a high degree of oral-based oriented English writing to a mixed orientation (e.g., González, Chen, & Sanchez, 2001;Rubin, Goodrum, & Hall, 1990), which appears to subscribe to the notions of L2 learners' idiosyncrasies and their language learning development.
The L2 learners' idiosyncrasies revealed from culture-specific rhetorical patterns of writing reflect what Selinker (1972) and Selinker, Swain, and Dumas (1975) have emphasized as an interlanguage, a learner-language system (Sampson & Richards, 1973) or an approximative system (Nemser, 1971). Based on Selinker, Swain, and Dumas's work on interlanguage (1975), the L2 learners may communicate using superordinate strategy called simplification. Simplification refers to the L2 learners' attempt to use their L1 translation equivalence in communication through two sub-strategies: language transfer and overgeneralization of L2 rules. Language transfer is defined as the learners' explicit applications of L1 rules to L2 norms. It means that the L2 learners construct English sentences or any parts of the sentences in the same way they would express the same meaning in their L1. Such transfer could be either positive or negative. Positive transfer usually facilitates L2 learners in the process of acquiring their L2 while negative transfer contributes some errors in the learners' communication messages. Overgeneralization of L2 rules refers to how L2 learners overextend some of L2 grammatical or semantic features in their sentences to a degree that those features do not apply. An example paragraph of the Thai EFL learner's simplification is illustrated in (1).
(1) "I think it's has a good taste…and it's cause of everyone use their package to take a photo to post on the internet and that's why it's..." (TEL10) In (1), the use of 'it's has…it's cause…it's…' displays apparent applications of overextending the use of copula be in the environment where non-copula is needed. Such applications may also represent the learner's use of 'it's' as the writer's proximal deictic expression which appears to be context-based or assume-known object of probably some food or drinks referring from the topic 'a good taste'. With the views of both L1 cultural and language interference, this high-context cultural writer may be perceived as inconsiderate by the readers who associate with the rhetorical and L2 norms of clarity and explicitness. Although (1) could also be broadly considered as an example of negative transfer, it could exemplify a developmental continuum where the learner moves from exhibiting negative transfer toward a more positive one in English communication (e.g., Derakhshan & Karimi, 2015;Selinker, 1983) through the learner's developmental and constructivist perspectives on discourse organizational patterns which involve syntactic-semantic-pragmatic means of English writing.

Decontextualizing L2 Learners' English Paragraph Writing Using Analysis Models of Metadiscourse Markers and Topical Progressions
A paragraph refers to one or a series of sentences which begin on a new line of a whole composition and is coherently organized as well as deals with a single topic of event, description, or idea. Through the view of this definition, paragraph writers go beyond grammar and vocabulary into linking linguistic elements to produce a unified, coherent whole composition at the discourse level. The way L2 learners organizes their texts also reflects interrelated language metafunctions which construe meaning ideationally, interpersonally, and textually (Halliday & Matthiessen, 2004). The L2 paragraph writers draw on ideational meaning or their exhibition of world experiences using their selective linguistic resources or textual meaning while managing to construct interpersonal meaning or their social roles and relationships with the readers. Thus, English paragraph writing could be considered as an appropriate source of authentic texts reflecting individual learners' real world of experiences.
In an attempt to comprehend such English texts with the interrelated language metafunctions, metadiscourse should be involved. It refers to a significant communication phenomenon where some particular linguistic features are employed to construct ongoing discourses, also known as "writing about writing" (Williams, 1981, p. 50); "communication about communication" (Vande Kopple, 1985, p. 83); or "discourse about evolving discourse" (Ädel, 2006, p. 2). Hyland and Tse (2004) proposed a metadiscourse model in which text organization and writer-reader interaction management are interwoven. The model constitutes two macro categories: interactive resources and interpersonal resources in which each governs five subcategories of discourse markers as illustrated in Table 2. The interactive category refers to the use of discourse markers under the five subcategories to guide the readers through the text. The five subcategories include transitions, frame markers, endophoric markers, evidential markers, and code glosses. Transitions connect ideas in the text through the use of linking adverbials based on their semantic types, such as addition, contrast, and consequence (e.g., Biber et al., 1999). Frame markers indicate boundaries of the text either as to sequencing, labeling stages, announcing goals, or shifting topic. While endophoric markers refer to other parts of the text by making additional material available for the readers, evidentials introduce sources of information from the other texts other than the current one. Code glosses are to expand the propositional meaning and reinstate ideational information.
The interactional category refers to the use of discourse markers under the five subcategories to involve the readers through the text's argument. The five subcategories include hedges, boosters, attitude markers, engagement markers, and self-mentions. While hedges indicate the writers' unwillingness to withhold commitment in presenting propositional information, boosters express the writers' certainty in doing so. Attitude markers display the writers' appraisal of propositional information. Engagement markers indicate the writers' attempts to make some relationship with the readers or address them explicitly. Self-mentions express the extent of the writers' presence through the use of personal deixis, such as first-person pronouns and possessives.
A closer look at the use of metadiscourse markers among L2 learners could give insights into continuum of rhetorical patterns of English paragraph writing among L2 learners from exhibiting a high degree of implicitness or indirectness through the use of hedges or non-linear order structure with the use of multiple additives, such as and toward a high degree of explicitness or clarity through the use of more sequencing transitions, frame markers, or boosters. Thus, it is possible that the writers whose L1 cultures echo oral-based rhetorical patterns of writing could employ more of additive transitions suggesting oral-based or reader-responsible writing orientation. On the contrary, the authors whose L1 cultures are close to those of literary-based ones could use more of more sequencing transitions, frame markers, or boosters suggesting literary-based or writer-responsible writing orientation. To further address how L2 learners achieve coherence in their writing from a more process oriented, topical structure analysis (TSA) or topical progression model should be highlighted. Lautamatti (1987) proposed TSA as to describe how L2 learners form coherence of their texts based on topics' sequences per sentence within a paragraph. The sequencing of topics moves through the text to progressively construct meanings the L2 learners want to communicate. In plotting the sequencing of topics, Schneider and Connor (1990) described three-step process involving: (1) identifying T-unit topics; (2) categorizing the progression of T-unit topics; and (3) coding the progression of the T-unit topics. In one T-unit topic, the three elements are taken into consideration. They include initial sentence element (ISE) or what come first in a sentence; grammatical subject/mood subject (GS) of a sentence; and topical subject (TS) or what the sentence is about.
By identifying T-unit topics, each clause consisting of ISE and GS or each subordinate clause is marked as one T-unit while coordinate clauses are counted as two T-units. When categorizing the progression of T-unit topics, the four types of TS development or progression (Lautamatti, 1987;Simpson, 2000) are considered as follows: A. parallel progression (PP) refers to two consecutive T-units with the same TS.
B. sequential progression (SP) is defined as the rheme element of the T-unit or the reminder of the message as the part in which the theme is developed (Halliday, 2014, p. 64) becoming the theme element of the consecutive T-unit.
C. extended parallel progression (EPP) refers to a TS that occurs in two T-units which are not consecutive.
D. extended sequential progression (ESP) is defined as the rheme element of a T-unit being taken up as a theme of a non-consecutive T-unit.
The four main types of TS progression could also occur in a sequential pattern, as Carreon (2006) proposed: By coding the progression of the T-unit topics as the four main types, the use of PP reflects the writers' repetitive structured texts where the given topics are constantly repeated as those displayed in the initial clauses' elements. When SP is selected, it displays the writers' constant and uniform discussions of related ideas. It also shows that the writers are fully aware of their English writing's content and a possibility in extending the content further. The use of EPP expresses a series of clauses whose first and final topics are the same with an interruption of SP. It means that the writers are able to structure their compositions back to the main theme. According to Carreon (2006), SP could be considered as a desirable topical progression of learners of L2 or English writing. The use of ESP could be viewed as similar to EPP. However, it highlights the writers' attempts to explore their given topics through linear explanation in a non-consecutive clause, taking the rheme of previous point of discussion as the theme which display a more complex and sophisticated return to a previous point of discussion when compared to EPP. The use of each type of the TS progression could suggest a continuum of rhetorical pattern of English writing among L2 learners from being simple deductive-oriented writers toward more complex deductive or writer-responsible oriented ones. Thus, it is possible that the writers whose L1 cultures echo oral-based rhetorical patterns of writing and whose English language experience remain low could employ more of PP and SP types of topical progressions suggesting simple deductive writing orientation. On the contrary, the authors whose L1 cultures are close to those of literary-based ones and whose English language experience are high could use more of EPP and ESP types of topical progressions suggesting a more complex deductive writing orientation.
Both metadiscourse and topical progression models appear to be in support of decontextualizing L2 learners' English paragraph writings. The models could allow us to systematically examine features of language and cultural interference based on theoretical grounds of interlanguage, L1 interference, and rhetorical interference reflecting interrelated language metafunctions which construe meaning ideationally, interpersonally, and textually (Halliday & Matthiessen, 2004).
Taking the views of communication in remote learning between the Thai EFL learners and their instructors in the Thai university context, interference and the learner-language system into accounts, this study used English paragraph writing as a medium of casual communication between the Thai instructor of English and the Thai EFL undergraduate learners during the second week of Cross-Cultural Contemporary course in an international program of a Thai university in January 2021 managed through Zoom platform. The 'getting to know me' activity was conducted to allow these Thai EFL learners to share their background knowledge on pop culture in writing and to post it via Padlet application similar to a bulletin board for online communicators. This writing activity was to accommodate the learners' communication preferences of writing rather than speaking, and to create a positive instructor-learner connection and relationship while encouraging learners' engagement in remote learning setting through their topics of interest relating to pop culture.

Research Questions
This study sought to answer the following two research questions: 1) What are the metadiscourse markers that the Thai EFL learners with high exposure to English language (TEHs) and those of the low exposure to English language (TELs) use to convey their pop culture knowledge in their English paragraph writing?
2) What are the types of topical progressions found in the English paragraph writing of the TEHs and that of the TELs?

Participants
93 Thai EFL learners who registered for Cross-Cultural Contemporary course in an international undergraduate program at one faculty in a major university in Bangkok, Thailand were invited to participate in this study after an information session at the first week of the semester in January 2021. This group of learners was in the same age group (17 to 20), sharing similar linguistic, cultural, and educational backgrounds, but with relatively difference in their experiences in learning English. All of them were asked to complete the English Language Exposure Questionnaire (Note 1). This questionnaire was selected as an alternative tool to a standardized English proficiency test in order to discriminate participants in the high English exposure group or TEHs and those in the low English exposure group or TELs (Note 2).

Procedure
Among the 93 Thai EFL learners, 15 participants were put in the TEH group. The other 15 participants were placed in the TEL group. The two sample groups were selected based on the distribution of their score results of the English Language Exposure Questionnaire, which revealed a significant difference among them (Duncan's, p < 0.05). The English exposure level of the TEHs was higher than that of the TELs. The higher English exposure level of the TEHs was from their higher scores in all three parts of the questionnaire, especially from partaking in more English self-practice activities and intensive English courses.
During the second week of the course in January 2021, the 'getting to know me' activity was conducted to allow the 93 Thai EFL learners to share their background knowledge on pop culture in English paragraph writing. They were asked to write about 'my kind of pop culture' between 100 and 150 words in length and to post their works via Padlet. Only the English texts from the two sample groups (N=30) were analyzed. The English texts from these two sample groups contained grammatical and diction errors. As to keep all the participants anonymous, the symbols of TEH1 to TEH15, and TEL1 to TEL15 were used to refer to each participant.

Data Analysis
The current study employed the metadiscourse and topical progression models for the investigation of the distribution and nature of metadiscourse markers and topical progression patterns in English paragraphs of the two sample groups. The models seemed to be suitable for the purpose of this study because they could reveal the way the TEHs and the TELs use the (a) interactive resources to make their texts more cohesive, (b) interactional resources to signal their interpersonal stance toward both the propositional content and the readers of the texts, and (c) topical progression to display the learners' overall writing organizations affected the comprehensibility of that knowledge on pop culture in such written forms. Instances of metadiscourse markers and topical progression patterns were manually coded and annotated for the frequency.
By coding metadiscourse markers, this study used the framework of metadiscourse markers developed by Hyland and Tse (2004) and Hyland (2005) as exhibited earlier in table 2. A coding example of transition subcategory under the interactive resource is displayed as in (2). Italics signal transition subcategory.
(2) "How the furby coming is Chompoo Araya have this toy and she was released on Instagram and her fan bought this toy and it was popular in Thailand." (TEL1) By coding the progression of the T-unit topics, two main procedures were followed. Firstly, it was to locate T-unit topics in a paragraph by taking into accounts the following three elements: initial sentence element (ISE), grammatical subject/mood subject (GS), and topical subject (TS). Italics were used for ISE which could be the subject of the sentence, and an introductory phrase or clause. Underlining was to signal GS which involves the mood or grammatical subject of the sentence. Bold face was for TS which may or may not be GS. Arrows were used to point to the development of topics. Some examples of T-unit topics are presented as in (3) and (4).
(3)   Vol. 14, No. 10; To achieve reliability in annotating the metadiscourse markers and topical progression patterns, the two reliability procedures were followed: (a) one fellow linguist in the field of discourse analysis' cross-checking the meaning of the codes (the inter-coding agreement reached 81%), and (b) revising the annotations.

Metadiscourse Markers of the TEHs and the TELs
As illustrated in table 3, the findings reveal the prominent macro category of metadiscourse markers under the interactive resources employed by both TEHs and TELs at approximately 80% to construct their pop culture knowledge in the English paragraphs. The interactional resources were used at approximately 20%. Among the 10 subcategories of metadiscourse markers, both groups of learners exhibited their preferences toward the use of transitions (57.39%), followed by frame markers (17.40%), hedges (8.68%), self-mentions (8.70%), code glosses (3.48%), engagement markers (1.74%), evidentials (1.74%), and boosters (0.87%). No frequency found in the use of endophoric markers and attitude markers. Although transitions were used at the highest frequency by both TEHs and TELs in their English paragraphs on pop culture, the TEHs employed more of them. The most frequent transitions used for both groups were the additive linking adverbial 'and'. Through the view of contrastive features of oral-based and literary-based rhetorical patterns of writing, the high frequency of 'and' used by both groups of learners could suggest the features of temporal structures between events (i.e., and…and…and then…and) (Van Dijk, 1973), of appositional connective 'and', and of narrating or dramatizing emphasizing the trains of thoughts similar to story-telling style. Through the use of these features, both sample groups' narrative style of writing appears to reflect one significant purpose of their English paragraphs which is to inform the readers. They tend to lead the readers to facts and words related to the pop culture topics of their interest.
Although both groups employ frame markers in similar frequency, the TEHs and the TELs tend to use the markers differently. The TEHs appear to prefer the use of finally, and for this reason while the TELs' preference is toward the expression "This is how I think." Such use among the TEHs could suggest a more structured text toward reader-focused feature or literary-based rhetorical pattern of writing. On the one hand, the TEHs tend to signal the readers the closure of the story and their emphasis on specific facts and words which allow them to justify their pop culture knowledge to the readers the way it is. On the other hand, the TELs tend to guide the readers based on their subjective inputs. Both groups of learners' composing tendency may reflect the authors' standpoints in persuading the readers in which the TEHs could be justified as having a more explicit and clearer degree of persuasion as compared to a more implicit one of the TELs.
In terms of self-mention markers, many TELs tend to employ the first pronoun 'I' as a self-mention interactional device while the TEHs prefer the pronoun 'We' which could be viewed as the speaker-reader oriented or reader inclusive pronoun when compared to 'I', the speaker-oriented pronoun. The TELs also used more code glosses, especially the use of such as. The higher frequency in the use of such code gloss marker among the TELs reflects their structured text with multiple aspects of topic. They tend to present the whole proposition and later extended to its part or related information. They also employed more hedges, especially 'I think' than the TEHs. The use of 'I think' among the TELs could display the writers' withdrawal from full commitment to proposition, which could be viewed as a feature of indirectness. It could also exhibit what Hyland (2010) proposed as 'proximity of commitment ' (p. 117) where the writers take their positions toward the issues in the texts. Hedges entail an attempt of both groups to take their positions although the TELs seem to exhibit a lower level of 'proximity of commitment' when compared to that among the TEHs.
As for endophoric markers, since the markers seem to function as to facilitate comprehension of the readers through referring to earlier propositional material and to anticipate other related material to come, none of the markers was used. This non-preference device appears to highlight inductive or reader-responsible oriented writing where the readers will actively seek the information and meaning themselves. In terms of attitude markers, since both groups of learners appear to reflect the main purpose in writing to inform the readers, it could be a possible reason of why attitude devices are not utilized in their paragraphs.
A closer look into the paragraphs of both groups of learners sees several ill matched clauses with patterns of expressions expected by readers in the target language.
(5) "How the furby coming is Chompoo Araya have this toy and..." (TEL1) In (5), the English T-unit that TEL1 composed may be perceived as deviated from patterns of expressions expected by the target language readers because of sentence fragment and an indirect question construction. One possibility of a well-formed sentence could be "How the furby is coming is that Chompoo Araya has this toy and…" It means that only a canonical or affirmative subject-verb agreement structure is required together with adding the relative marker that to refer to the antecedent act and allowing have to agree with its subject Chompoo Araya. Apart from these deviations, (5) could be perceived as to display the learner's attempt in calling readers' attention. Instead of forming the sentence as 'How is the furby coming?', the learner constructs a deviation form of the declarative sentence 'How the furby coming is…'. Interestingly, this deviant construction could allow an interpretation of the verb 'come' that the learner used as to denote the concept of result (Sutthichatchawanwong & Luksaneeyanawin, 2006) and not that of direction. In addition, the deviant form of the verb 'come' as 'coming', has a present progressive marker suggesting that the event is in progress at a given moment. Although to some extent such deviations could be perceived as illustrating L1 negative transfer when taking Thai translation equivalence into account, to the other extent the form reflects the notions of hypercorrection of the sentence. Rubin (1987) and Rubin, Goodrum, and Hall (1990) suggest that hypercorrecting writers weigh down their syntax and select indirect expressions. This example of hypercorrection could exemplify the strategy involving the readers in the text or as using interactional marker of engagement. TEL1 formed the mentioned English T-unit as an implicit engagement marker to build relationship with readers. It could be the reason why explicit engagement markers this study offered to highlight in analysis did not occur frequently among the TELs. It is also observed that the similar composition as TEL1 wrote appears among some TELs, but not among the TEHs.
(6) "Totally it have at least 3,000 branches all over Thailand." (TEH10) In (6), the English T-unit that TEH10 structured could be viewed as exhibiting subject-verb agreement violation. Although this rhetorical pattern of expression may not pose a high degree of interference toward the target language readers' comprehension, it displayed the learner's L1 negative transfer of an absence of subject-verb agreement (between the subject 'it' and verb to 'have') from Thai to English. While some TEHs exhibited this characteristic of L1 negative transfer, all TELs express it. Apart from the language interference, a closer look at the TEH10's English T-unit found a distinct characteristic which reflects the use of booster 'totally', an interactional marker to emphasize the writer's certainty in the proposition. Although the occurrence of such use is relatively low, there is none in the TELs' writing. Exhibiting certainty of TEH10 could exemplify the English high-exposure learner's development which moves closer toward the patterns of expression with clarity and explicitness expected by the target English language readers.

Types of Topical Progressions of the TEHs and the TELs
It appears that the TEHs constructed more independent clauses and new topics in their compositions than did the TELs. This could derive from the TEH's ability to construct more words in English (approximately 1,725 words) as compared to that of the TELs (approximately 1,200 words). Overall, both groups of learners structured their English paragraphs using the three types of topical progressions: PP, SP, and EPP. None of ESP was used. The TEHs appear to employ more frequency in each of the three types as shown in table 4. Apparently, when writing their English paragraphs on pop culture, both TEHs and TELs exhibit their preferences in simple deductive-oriented progression. The TEHs and the TELs employed more SP in their texts, followed by PP and EPP. The TEHs tend to show a relatively high degree in constructing the three types of topic progressions in their English paragraphs as compared to that of the TELs. The use of SP in both groups of learners' texts could be viewed as the learners display their constant and uniform discussions of related ideas, which reflect a valued characteristic of English composition as found in previous studies (Carreon, 2006;Flores & Yin, 2015;Simpson, 2000). Furthermore, it could be perceived as the learners are fully aware of their English writing's content and a possibility in extending the content further into a more complex style of writing, such as an argumentative essay. The use of PP could reflect the learners' repetitive structured texts where the given topics are constantly repeated as those displayed in the initial clauses' elements. Unlike the previous studies which the use of PP displays the writers' repetitive structured texts (e.g., Carreon, 2006;Simpson, 2000), the use of PP revealed in this study appears to exemplify the learners' repetitive set of lexicons, especially among the TELs (see figure 2). This could explain the reason why the TELs create fewer independent clauses and new topics in their compositions. It is not surprising that ESP was not found in the data. The use of ESP highlights the writers' attempts to explore their given topics through linear explanation in a non-consecutive clause, taking the rheme of previous point of discussion as the theme. Thus, constructing ESP in the learners' English paragraphs display a more complex and sophisticated return to a previous point of discussion when compared to EPP. Overall, the preference of both groups of learners toward SP, PP, and EPP types of topical progressions suggesting their writing development from a simple deductive writing orientation toward a more complex deductive and writer-responsible writing orientation.

Conclusion
This study investigated features of interference in the Thai EFL learners' English paragraphs on pop culture. Drawing from theoretical grounds of interlanguage, language interference, and rhetorical interference with the analyses of metadiscourse markers and topical progressions, the Thai EFL learners' writing on pop culture exhibits context-based, concrete information, multiple aspects in focus, additive, narrative and reader-responsible English texts which may deviate from the expectation of the target language readers of text-based, single aspect in focus, and writer-responsible English paragraph.
The findings reveal similarities and differences between the TEHs and TELs in organizing their English paragraphs. In the similar vein, the high frequency of 'and' used by both groups of learners could suggest the features of story-telling style in writing, speech-like writing, or additive-narrative style of writing which reflect informative purpose of their idiosyncratic English paragraphs. The difference in marking the writers' frames as to help guide the readers through the texts lies in the use of objective reasoning (for this reason) among the TEHs as compared to that of the subjective one (This is how I think) among the TELs. In involving the readers through the texts by using some interactional devices, many TELs tend to employ the first pronoun 'I' as a self-mention marker while the TEHs prefer the pronoun 'We' which could be viewed as the speaker-reader oriented or reader inclusive pronoun when compared to 'I', the speaker-oriented pronoun. The TELs also used more code glosses, especially the use of such as which could be perceived as to reflect their structured texts with multiple aspects of topic. They also employed more hedges, especially 'I think' than the TEHs. The use of 'I think' among the TELs could display the writers' withdrawal from full commitment to proposition, which could be viewed as a feature of indirectness. It could also exhibit what Hyland (2010) proposed as 'proximity of commitment' (p. 117) where the writers take their positions toward the issues in the texts. Hedges entail an attempt of both groups to take their positions although the TELs seem to exhibit a lower level of 'proximity of commitment' when compared to that among the TEHs. Both groups of learners' preferences in the use of interactive and interactional devices as well as of SP, PP, and EPP types of topical progressions highlight features of oral-based, inductive, or reader-responsible writing orientation with a possibility of writing development, especially among the TEHs to reach expectation of the target language's readers.
As emerged in this study, each group of learners displayed the common interests in music, celebrities, Netflix series, movies, video games, food/drinks, and fashion which suggest contemporary forms of pop culture the learners have experienced. Their common ground knowledge of these topics in pop culture tends to be US-South Korea centric exemplifying dominant power dynamics of current global media culture as well as their personal interests. These learners' interests could further assist in planning a more meaningful content for the Cross-Cultural Contemporary course and other English lecture courses dealing with cultural learning. One possible option is to use a range of Netflix series, movies, or celebrities as exemplars to relate pop culture to globalization, identity, health care, or genders and encourage more informal English writing activities with feedbacks from the readers. Although the findings strengthen previous contrastive studies of high-context as opposed to low-context cultural writings, they suggest focusing on the learners' English writing as a process and not solely as a final product. In the case of the TELs, they need more input of English lexicons to be able to extend their paragraph writing. Instructors may assign more extensive reading to help building more vocabulary for the learners (e.g., Liu & Zhang, 2018). In addition, the deviations in the writings of the TEHs and the TELs suggest the need to encourage more use of deductive-oriented types of progressions for an increased sophistication of English paragraphs. In so doing, it could assist them in becoming aware of language and rhetorical interferences in writing English texts.
As the introductory stage during COVID-19 remote learning, integrating the real-world experience of the learners as with the case of pop culture in this study suggests a way to build a foundation for a positive instructor-learner relationship which assists the learners to further engage in the class in a more meaningful way during such learning context. It may also help strengthen evaluations of remote learning courses' context, input, and process for an effective online learning.