Exploring Students’ Perception of Subject and Translation Integrated Learning (STIL) as an Innovative Approach to Translation Instruction: A Case Study

Subject knowledge, also known as domain knowledge or thematic knowledge in translation studies (TS), is universally recognized as a constituent element of translator competence. However, there is a dearth of empirical exploration into its instruction in translation classroom, especially from translation students’ perspective. Inspired by CLIL (content-language integrated learning) in L2 learning and relevant discussions in TS, this study proposes STIL (subject-translation integrated learning) as a novel approach to translation instruction characterized by integration of a systematic in-depth instruction of subject knowledge with translation instruction on course level. It then explores students’ perception of the proposed approach as it is implemented in an MTI (Master of Translation and Interpreting) course specialized in translation of classical texts in economics. The results show: (1) STIL is generally preferred by the participants to other approaches to translation instruction that they have been exposed to; (2) STIL, if well-designed and carried out, could produce positive cognitive and affective outcomes with both subject-learning and translation-learning through a forced and facilitated engagement with the texts.


Introduction
The importance of subject knowledge has been welldocumented in the history of TS.As early as in 1540, Dolet (1509-1546)), generally considered as the first theorist of translation, emphasized that the translator must understand the contents of the source text (as quoted in Gonza´lez Davies, 2004).This recognition in TS of content knowledge as a necessary part of translators' knowledge repertoire has been further endorsed by contemporary translation scholars under the name of subject knowledge, domain knowledge or thematic knowledge (Salamah, 2021).Neubert (2000, p. 4) states that '' Translation, by definition, introduces new knowledge'' and translators therefore must possess a certain level of knowledge of the subject involved sufficient for its proper rendering in a new language.Shreve (2006) claims that experts exhibit greater domain-specific problem awareness than non-experts and develop conceptual representations differently from beginners, allowing them to retrieve task-relevant information and knowledge to solve problems more efficiently and effectively.Sharkas (2013) finds out that those reading texts of relevant science knowledge before translation produce a better translation than those without such reading experience.Given the proposals and evidences of this sort, subject knowledge naturally finds its way into almost all modern models of common translation competence (EMT, 2017;EMT Expert Group, 2009;Go¨pferich, 2009;PACTE 2003PACTE , 2017) ) and is highlighted as a key sub-competence for specialized translations such as for legal translation (Parra-Galiano, 2021;Prieto Ramos, 2011;Stolze, 2011) Despite this universal acknowledgment of the role of subject knowledge, few inquiries, have ever been made to explore the effective ways to acquire or teach it in translation classroom.Of the few, Anderman and Rogers (2000, p. 67) reviewed two approaches to subject knowledge instruction in UK translation programs: the centripetal one and the centrifugal one.The former, characterized by an integrated course of lectures specially designed for student translators, has been widely adopted in specialized translation programs or those aimed to bring up so called inter-disciplinary translation talents, where a large proportion of subject courses are offered in parallel with language and translationfocused courses.The major merit of this approach consists in its readiness of implementation (Prieto Ramos, 2011;S ˇarcˇevic´, 1997).Although it is easy to implement, its effectiveness is highly doubtful.Prieto Ramos (2011) maintained that this approach can not ensure sound decision-making in translation and Kelly (2007) criticized it for resulting in poor ability of students to relate the various components of a curriculum to one another in ''a sadly impermeable set of separate compartments of knowledge' ' (p. 138).In other words, this centripetal approach may fail to help students establish an intimate relevance of the subject knowledge delivered in lectures and the translation practices in which they are involved.Moreover, with limited program length split up between the independent sets of courses in this approach, many graduates left the program with a poor command of both subject knowledge and other subcompetences of translation (X.Shen. personal communication, May 12, 2021) The latter is a pure text-based approach in which no subject lectures are offered and translation of text is considered as a kind of window on related and relevant skills, principles and knowledge.This centrifugal approach is said to have the advantage of exposing students to a wide range of subject fields.Yet, the subject knowledge acquired in this manner for each subject may lack the systematic depth required of quality translation (Gonza´lez Davies, 2004;Prieto Ramos, 2011).On the one hand, being simply exposed to more subjects means less time spent on each of them.On the other hand, it is hard for the short texts chosen for translation to represent the subject knowledge in its proper width and depth which documents to be translated are supposed to be situated in (Prieto Ramos, 2011).In contrast with dealing with many subjects, there are some strong advocates for shrinking the scope of subjects taught in order to introduce students to the principles of researching a subject and offer a template for methods and principles with which to approach other subjects (as quoted in Plein,1989).This advocacy suggests the importance of acquiring in-depth subject knowledge for translators.Yet as for how each of the subjects shall be taught, there seems to have been little exploration.
In short, it is agreed in TS that translators are required to acquire in-depth subject knowledge.However, with obvious weaknesses with each of the existing approaches as discussed above, there seems to be no agreement regarding how to equip students with such knowledge and know-how in translation program.Moreover, the discussion regarding advantages and disadvantages appear to have no support of students' views based on empirical data.To address this situation, STIL, an innovative approach to translation instruction inspired by CLIL, is proposed to help learners acquire knowledge in both economics and translation in an MIT course entitled Translation of Classical Texts in Economics, and its impacts as perceived by them are explored.In the rest of the paper, we will (1) review briefly CLIL as a successful approach to language teaching, (2) examine the rationale of STIL in translation instruction, (3) describe the implementation plan of STIL in this study, (4) report the research methodology of this study, (5) present and discuss the learners' perception of STIL, and (6) deal with the implications and conclusion of this case study.

A Brief Review of CLIL
CLIL refers to ''a dual-focused educational approach in which an additional language is used for the learning and teaching of both content and language'' (Coyle et al., 2010, p. 1).
In fulfilling these two aims, CLIL has adopted diverse implementation plans.For example, Coyle (2005) divided the various CLIL practices into five models as below: (1) Subject/topic syllabus, which involves teaching a subject in the target language to explore the subject from a different perspective.
(2) Cross curricular projects, where both language teachers and subject teachers plan together to teach different aspects of the same topic.(3) Content/theme-based approach, which is characterized by a comparative study of a theme investigating differences regarding a given concept in different countries/cultures.(4) Integrated curriculum in which the learners may study a concept in a foreign language from different perspectives.(5) Project-based approach where identical topics are studied by learners in different countries and the outcomes of the research projects can be compared by the learners.
In addition, Ball (2009) represented the models of STIL in a continuum as regards the changing degree of emphasis given on content-learning with the opposing ends labeled as ''strong version'' and ''weak version'' as shown in Figure 1.
It is supposed that all these different models embody the flexibility of CLIL in adapting to different contextual demands such as the institutional expectations, curriculum options, teachers' pedagogical orientations, learners' age and language levels, and so forth.
Regardless of these variations in practice, the general empirical findings so far have indicated that CLIL, if properly implemented in its particular context, can yield positive learning outcomes in cognition and affection.Cognitively, it has produced superior learning gains in language ability, subject knowledge, intercultural knowledge, diverse learning strategies, general cognitive and thinking skills, and so on (Badinska, 2010;Bulte´et al., 2022;Jafarigohar et al., 2022;Mo¨ller, 2017;Orozco & Pedrosa, 2022;Rose Mahan & Norheim, 2021).Affectively, it has resulted in learners' better attitudes toward language, higher motivation for learning and reduced learning anxiety (Badinska, 2010;Bulte´et al., 2022;Lasagabaster, 2009Lasagabaster, , 2019;;Rose Mahan & Norheim, 2021;San Isidro & Lasagabaste, 2022).Thanks to these proven benefits, CLIL is often considered as an efficient way to teach subject content and language together (Tedick, 2020) and has become a widely accepted educational approach in the domain of foreign language education across the world (Hemmi & Banegas, 2021).
Despite its widespread success, CLIL is thought liable to several challenges, namely, reduced coverage of subject matter in less complexity resulted from use of L2 as instruction language, the professional rivalry between different groups of teachers with some teachers feeling guilty for ''acting too much like a language teacher'' and undermined learners' ability to express content knowledge in L1 (Dalton-Puffer, 2007;Karabassova, 2020;Metli & Akis, 2022;Rose Mahan & Norheim, 2021).

Rationale of STIL
In light of the impressive achievements of CLIL in the face of challenges as mentioned above, STIL is supposed to be at least equally successful on the grounds below.

Translation as Part of Language Learning and Teaching
Language learning and translation learning have been naturally intertwined since the time when grammar was the main part of language pedagogy.Now in the era of communicative language teaching, translation, which requires a deep processing of linguistic signs and meets almost all usual standards of TBLT (task-based language teaching), is considered as a justified means of language instruction (Carreres, 2014;Cook, 2010).As a matter of fact, the integration of translation in language teaching seems to have gained steady ground with more and more theoretically-informed creative ways explored and implemented to introduce translation into language teaching in this age characterized by increasing advocacy of bi-, multi-, or plurilingualism and international mobility (Carreres et al., 2018(Carreres et al., , 2021;;Gonza´lez-Davies, 2018, 2020, 2021;Pintado Gutie´rrez, 2018, 2020, 2022).In short, language learning nowadays must and shall preferably involve translation in some ways and to some extent.It is in this manner that STIL may be regarded as a special variant of CLIL.

Inextricable Link Between Content Learning and Translation Learning
The marriage between subject learning and translation learning seems to be more solid and substantial than that between content learning and language learning.
For one thing, subject learning is more relevant to translation learning than content learning to language learning.While content knowledge is rarely stressed as part of linguistic ability, subject knowledge, as observed above, has long been acknowledged as part of translation competence.Whatever is done to promote subject knowledge will reasonably benefit translation learning.Firstly, subject knowledge can enhance the quality of translation products (Gonza´lez Davies, 2004).Secondly, subject knowledge gained can excite positive emotional experiences in translation process.For example, Prieto Ramos (2011, p. 13) asserted that: ''The deeper the knowledge of legal subjects, the more confident the translator can feel when dealing with legal content issues during analysis and transfer stages of translation.''In this light, the time spent on exploring subject knowledge, instead of taking the time away from translation learning, will probably benefit both the process and product of translation.
For another, what is invested into translation learning is supposed to promote subject learning, too.Translation is universally acknowledged as the most comprehensive act of reading (Morel, 2006) and translators as the most intimate readers of literary or scholarly texts (Porter, 2013), since they have to take account of all of the subtleties of the text in the process of translation.Hubscher-Davidson (2018, p. 3) pointed out: Translators undertake very close readings of source texts and become involved with source authors and their texts to a greater extent than do ''regular'' target readers.It can be said that the translation process binds them to source authors, providing them with intimate knowledge of how they work, how they construct meanings, impart knowledge and express themselves.
The ''closer reading'' of ST and the involvement with its authors make translation a deep learning process of subject knowledge.According to Hills (2012), translating emotional material forces translators to explore emotions and language on a deeper level, intensely, to find ways to communicate what is core in a piece.Likewise, translating subject material may deepen their comprehension of the subject by compelling them to examine the subject knowledge and its carrier in a more profound and intensive way.In addition to engagement with ST, translation also involves a close reading of TTs with translators drafting and re-drafting their own TT and reading TTs by others in case of re-translation.Such engagement with TTs will add to translators' comprehension of ideas carried in the texts.Firstly, students benefit from all their linguistic resources for knowledge construction and deep learning by being confronted with input in both languages (Baker, 2003).Secondly, they are exposed to different translators' individual interpretations of the same content.Humboldt (1963, p. 97) claimed that ''part of the nation that can't read the ancient classics for themselves can become better acquainted with them through several translations than through just one'' because different translations represent different views of the same spirit from different translators.In short, translation learning is expected to promote subject learning via a special engagement with both ST and TT(s).However, the discourse in this respect is mostly confined to literary works with specialized or academic texts left almost unexplored.

Immunity of STIL to the Challenges of CLIL
STIL may be more feasible than CLIL for it is free of the major inconveniences that might hamper CLIL.Firstly, both translation teachers and learners are supposed to have reasonable command of source language and target language, both of which can be used in a translation classroom, thus eliminating the chance of reduced depth and coverage of subject matter that might result from inadequate proficiency of foreign language.Secondly, the rivalry between subject teacher and translation teacher is less likely to occur as subject knowledge and translation are mutually supportive of each other and translation teachers are expected to have sufficient subject knowledge and take charge of both subject instruction and translation instruction at the same time.Thirdly, translation, which explores the link between ST and TT, will promote learners' L1 instead of undermining it.
Given these considerations above, STIL, if properly implemented, will reasonably bring about at least as many positive outcomes to learners as CLIL does.

Implementation Plan of STIL
For proper implementation of STIL, an instruction plan is designed for an MTI course entitled Translation of Classical Texts in Economics on the principle of optimizing subject learning and translation learning simultaneously.The particular practice and its rationale are as below.

Selection of WN as the Texts
Wealth of Nations (WN) and its multiple translation versions are chosen as the texts of instructional material.Such selection is motivated by the recognition that they are carriers of rich knowledge in both economics and translation and therefore of great value in fulfilling the dual-focus course objectives: subject learning and translation learning.
As the seminal work for economics as a discipline, WN is hailed for being a clear, concise and systematic synthesis of what past work had seen only through a glass darkly (Spencer, 2012).Rosenberg (1979) characterized the history of economics over the past 200 years as a non-stop and on-going annotation of WN.Endorsed in this manner, WN is believed to be a trustworthy source of rich subject knowledge in economics.
Beside its contribution to economics, WN is also considered classical for its language.Skinner (1986) believes that it is the weapons of rhetoric and irony as well as the telling force found in the volume that helps catch and hold the attention of its readers.Spencer (2012) thought what makes the books enjoyable for his contemporaries and ours and both academics and general readers alike are Smith's poignant examples and maxims, realistic description of the pin factory, the elegant simplicity of Smith's phrases, quotable lines that stick in one's mind, the power of his adages and his abundant illustrations drawn from the nooks and crannies of daily life.To translate ST of these features into TT must require a rich repertoire of transfer skill and strategies.Moreover, there are a huge number of Chinese translation versions of WN, which can serve as valuable materials for translation learning.For one thing, different versions of translations, which reflect translators' subjective differences in dealing with the texts, can be of great pedagogical value (Reiss, 2014).For another, the existing translation versions, which are of varying qualities, offer a huge collection of examples of translation errors, losses and creativity that arise in the process of linguistic transfer.Therefore, WN, together with its translations, is hopefully a quality carrier of translation knowledge.
In short, the original ST and TTs of WN are supposed to be a rich mine of knowledge in both economics and translation to be explored in translation classroom.

Development of Tasks Based on the Texts
In order to tap into the rich mine, two branches of activities have been finalized based on over 5 years of experiences with using WN and its translations as teaching materials: subject-focused activities and translationfocused activities.The objective of the first branch is to help the learners acquire subject knowledge of the texts chosen systematically and appreciate its relevance to their present life and progress of the discipline while that of the second to assist them in gaining translation knowhow.
For subject-focused activities, three sets of questions are designed by the teacher for each unit composed of several chapters working on the same theme: theory comprehension questions (TCQs), theory application questions (TAQs), and theory expansion questions (TEQs).For example, unit 1 is composed of the first three chapters of Book I, WN, which looks at the same topic of division of labor from different perspectives and for this unit, eight questions are formulated respectively under the category of TCQs, TAQs, and TEQs, as shown in Table 1.
Of these three sets of questions, TCQs attempt to give the learners the main ideas of each unit; TAQs to inspire them to use the key ideas in each unit for explanation of what is going on in their life; and TEQs to encourage them to connect the ideas in each unit with theories of economics after WN.For example, the three TCQs in Table 1 are expected to acquaint them with the effect, cause, and extent of division of labor; the two TAQs to encourage them to use the notion of extension of labor division and market to explain social and economic practices around them; and the three TEQs to induce them to compare Adam Smith's account of division of labor with the theory of comparative advantages by David Ricardo and that of utility maximization by the Chicago School.
In order to answer them, they are required to read unit chapters in both ST and TTs before class and discuss their answers in class with the teacher who shares his own answers with them too.It is hoped that these activities will be able to equip them with not only adequate subject knowledge that will make them cognitively and emotionally ready for subsequent translation-focused activities and but also with a model of learning classical texts.
For translation-focused activities, three tasks are developed based on ST and TTs of the same unit: Translation Error Task (TET), Translation Loss Task (TLT), and Translation Creativity Task (TCT).Translation errors refer to the indisputable mistakes with TT resulting from translators' poor understanding of grammar or negligence and the ability to identify them constitutes a basic aspect of translation expertise.Translation loss happens when TT fails to carry something in ST.Translation is known as an art of loss and an insight into the losses and ways for making up for them is highly expected of professional translators.Translation creativity signifies the quality or property of TT being close to ST in meaning and distant from it in form and the capacity of finding creative solutions is regarded as an essential indicator of expert competence.To complete these three tasks, the learners are required to read ST against TTs.For TET and TLT, they shall: (1) identify as many errors and losses as possible to the best of their knowledge; (2) share their own opinions and solutions within their respective teams; (3) present as a team what, in their view, are worth reporting in classroom.For TCT, they shall identify creative TT in existing TTs and/or produce their own creative versions and share the creative TTs either found or produced by themselves with their team members and the whole class in the same way with TET and TLT.The teacher will provide feedback on their performance and share his own work with them.

Course Composition and Time Allocation
The course consists of 5 units dealing respectively with five topics of economics explored in different chapters of

Activity type
Learning questions

TCQs
(1) What are the effects of division of labor?
(2) What is the cause of division of labor?
(3) By what is the extent of division of labor determined?TAQs (1) How do you explain the slogan ''In order to become rich, build road'' (2) How would Adam Smith think of the role of e-commerce in production of wealth?TEQs (1) Can division of labor cause any negative effects? (2) How can labor be divided to maximize the produce of labor?(3) What do you think of ''It is not from the benevolence of the butcher, the brewer, or the baker, that we expect our dinner, but from their regard to their own interest''? WN: (1) division of labor; (2) origin and use of money; (3) price; (4) wage, profit and rent; (5) capital.These topics cover the basic and essential knowledge in the domain of economics and the chapters focused on them in WN are supposed to offer a comprehensive and insightful analysis of them.As for time allocation, each unit is composed of three sessions (1.5 classroom hours for each session) with the first session assigned to subject learning and the last two to translation learning.Evidently, more time is set for translation-focused activities than subject-focused activities while an equally important dual goals are proposed for this course.The seeming disadvantage of the latter in time assignment can be compensated by reference to and discussion of subject knowledge which will naturally happen in translation-focused activities presented above.
In short, it is expected that by going through the two branches of activities as described above, learners will acquire both subject knowledge and translation knowledge effectively and efficiently while developing positive attitude toward them.

Research Design
While it is hypothesized that STIL, if implemented as above, will produce positive outcomes with both subject learning and translation learning, it remains unclear how learners will respond to it.The overarching aim of this study is to explore students' perceptions of STIL instruction, which can be reliable and predictive of learning (Wallace et al., 2016).

Research Questions
To fulfill this aim, two research questions are formulated as below.
RQ1: How do learners like STIL as implemented as opposed to other approaches to translation instruction they have experienced?RQ2: How do learners perceive the role of STIL as implemented in both subject learning and translation learning?
The answers to these two questions can help stakeholders understand how students perceive STIL teaching and how it can be improved.

Participants
The 14 participants of this case study are from an MTI program run by a university in China.As the university's strength lies in instruction and research in finance and economics, the overall objective of its MTI program is to equip its students with both expertise in translation and solid knowledge in finance and economics with the view to empowering their professional development as translators in the aforementioned domain.The course, Translation of Classical Texts in Economics, as part of the program, is aimed at educating students with basic principles of economics and knowledge of translation simultaneously and effectively.All the participants took the course during the first semester of academic year 2020 to 2021 and they acknowledged that they had taken in their undergraduate programs and/or were taking in their current MTI program translation courses in which subject knowledge was either completely ignored or handled lightly and causally in the process of translation instruction.

Researcher Positioning
The researcher of this study and the instructor of this course is the same person.In the role of the instructor, the researcher is quite familiar with the course details and participants and went through the whole learning process as the participants did, which enabled him to have a partial insider (emic) perspective and served as the researcher-as-befriender (Sarangi & Candlin, 2003).As both instructor and friend, the researcher was in a better position to encourage the participants to be both honest and thoughtful in their confessions.Moreover, as part of the efforts to stay free of the influence of his possible subjective preference for STIL on the participants, the researcher kept reminding himself of being objective in the whole process of research.

Data collection and analysis
To obtain rich and trustworthy insights into the topic, multiple instruments of data collection were employed as shown in Table 2.
The data were collected and processed through content analysis, a type of qualitative research that attempts to identify patterns, themes, and meanings through Semi-structured with SCI and TR as stimuli systematic examination of a body of material (Berg, 2009).I resorted to conventional content analysis, deriving codes inductively from the data itself using a grounded approach (Berg, 2009).It is hoped that in this manner we can find coherence in the findings to formulate common understandings of the students' perceptions of STIL.The codes were grouped together into common themes to find patterns.Each statement was categorized under a theme, and was given a numerical value to show how large each theme was.All these analyses were reviewed and verified by a colleague qualitative researcher to minimize the possible impacts of researcher's subjectivity.

Research Results and Discussion
The data from the various sources establish that STIL is preferable to students for its positive impacts on subjectlearning and translation-learning as illustrated in the extracts below.
Extract 1: STIL is more comprehensive.In STIL class, we are given the chance to understand the subject deeply and improve the translation while in other class, the focus is put on translation alone.I am interested in finance and economics and I had attempted to read relevant classics.But the progress was quite slow and I failed to complete them.In this course, I acquired a new way of reading: to read by comparing ST and TT or analyzing TTs and to deepen the comprehension of ST for the purpose of improving translation quality.(SCI, S6) Extract 2: I strongly endorse STIL for its focus on both subject-learning and translation-learning with WN and its multiple translations as the learning material.The work on errors, losses, and creativity in translations is itself an ongoing process of applying subject knowledge to grasping thoughts of Adam Smith.In other courses, subject knowledge is scarcely dealt with and when it is, it is usually like a dragonfly skimming water surface without in-depth exploration.What is more, the materials used are quite simple such as those from The Economist, carrying no subject knowledge as profound and systematic as found in WN. (SRI, S1) Extract 3: In the subject-learning part, the teacher will walk us through the text by discussing the questions designed for theory comprehension, application and expansion, which gives me a deeper and clearer idea of the key subject of ST and makes the subsequent translation tasks easier.

Results and Discussions Regarding RQ1
The SCI data show that translation instruction can be roughly divided into two approaches: STIL and non-STIL where subject instruction is either ignored or downplayed (Extracts 1 and 2).And of the two, 13 of all the 14 participants stated explicitly their preference for STIL mainly because it stands for a new approach that has effectively produced learning gains both in subject and translation, which will be further examined in 6.2.The one who did not make such a statement acknowledged that STIL is better than non-STIL in developing her understanding of the subject (Extract 1).

Results and Discussions Regarding RQ2
As illustrated in the extracts above, STIL seems to have played a positive role in subject learning and translation learning on both cognitive and affective dimensions.
The Role of STIL in Subject Learning.As far as subject learning is concerned, the most salient cognitive outcome is the depth of subject knowledge acquired as indicated by such expressions as ''deeply'', ''in-depth'', and ''on a deep logic level'' (Extracts 1-5).Further analysis shows that such profound comprehension of subject knowledge can be ascribed to three circumstances: (1) texts rich in subject knowledge; (2) subject-focused activities; and (3) subsequent translation-focused activities.Texts, as carrier of subject knowledge, differ in extent and depth of subject knowledge.WN is perceived by learners to be abundant in knowledge of economics (Extracts 2 and 6) and the subject-focused activities developed around them are universally approved by all the participants for its role in enhancing their subject knowledge (Extracts 3 and 4).The subject knowledge thus gained is then deepened and internalized by carrying out the structured translationfocused activities (Extracts 1, 2, 6, 7, and 13).The way that translation-focused part contributes to acquisition of subject knowledge provides from students' perspective a further support to the established view that translation represents a comprehensive, close, and critical form of reading (Hubscher-Davidson, 2018;Morel, 2006;Porter, 2013) and an effective approach to deep-learning of subject knowledge and suggests that the effect could be magnified by well-designed engagement with the subject in both ST and various TTs.
The most salient affective outcome for subject learning seems to be the change of learners' attitude toward classical texts or WN in particular, resulted from their (1) increased awareness of the value of ST in question and (2) boosted confidence in deciphering it (Extracts 4, 9-13).In addition to the subject-focused part that opens learners' eyes to the value of ST by acquainting them with its basic ideas and relevance (Extracts 4 and 11), the translationfocused part also plays a significant role in cultivating their attitude toward classical texts.Firstly, it serves as a productive assignment which forces the participants to engage with the classical texts in order to produce tangible translation criticism and products.In other words, for the sake of completing the translation tasks, learners have to peruse the demanding classical texts in both ST and TT.Without this specific product-oriented translation task, learners would easily give up studying these tough texts as they did in regular reading attempts for general comprehension (Extracts 1 and 12).Secondly, the translation learning part facilitates the digestion of the classical texts as a special way of studying them (Extracts 1 and 13).In this process of close and deep interaction with the classical texts, the participants have come to realize that classics are not as inaccessible as they used to think while being made aware of their value in both content and form (Extracts 8,9,10,and 11).The recognized value of ST can further motivate them to explore the ST and its subject knowledge.In this light, the translationfocused part, can be considered as a forced and facilitated process of building up knowledge of and attachment to ST interactively.
The Role of STIL in Translation Learning.As for translation learning, the obvious cognitive outcomes include: (1) the awareness of the role of deep subject knowledge in translation learning; (2) the improved command of learners' L1 and (3) the recognition of the value of classics for translation learning.It is common knowledge that subject knowledge is necessary for successful translation.However, as a result of experiences with the STIL, the participants have been made conscious of the impact of the degree of subject knowledge comprehension on quality of translation.This echoes the belief in the context of legal translation that a deeper knowledge of law makes a better law translator (Prieto Ramos, 2011).A closer look into the data reveals that the subject-focused learning can facilitate translation learning by both reducing learning difficulty and improving translation performance.Classical academic texts are hard to read, not to speak of translating.That may be partly why these texts of many merits are rarely used for translation instruction.The subject-focused part, as expected and confirmed by the participants, reduces to some extent their difficulty, paving the way for the subsequent translation-focused activities (Extract 3).Specifically, the subject knowledge gained has helped the learners identify the errors and losses in existing translations (Extract 15) and produce creative translation solutions (Extract 14).These reports highlight translators' command of deep subject knowledge as an essential player in performing these tasks.In addition to the subject-focused part, translation-focused part contributes to learners' translation performance too by improving learners' L1 (Extract 7), as opposed to L1 being undermined in CLIL (Rose Mahan & Norheim, 2021).Last but not least, thanks to STIL carried out in this study, the learners have come to realize that, classics, though difficult, could be good materials for translation learning if well-developed (Extracts 12 and 13).
In addition to cognitive gains, STIL seems to have produced desirable affective outcomes for translationlearning in terms of: (1) sense of responsibility of translators in transmission of valued human knowledge; (2) fun and confidence involved in creative renderings; and (3) motivation to become a better translator.Specifically, STIL, with both subject-learning part and translationlearning part combined together, supports and sustains learners' engagement with the classical work in both ST and TTs.Such an engagement has led to their deeper recognition of both the great value of ST and the poor quality of existing TTs, which has created in them a clearer awareness of critical role of translators in decent rendering of classical ST (Extract 9) and a stronger motivation to make a better one (Extract 16).In the context where translators and translations are regarded as secondary to writers and originals, this heightened sense of value of translators and intrinsic desire to improve their performance is essential for successful translation practice and education.In the attempt to produce a better translation, they find themselves in want of necessary knowledge and resources for producing a classical TT, which drives them to work harder on improving their translation product and themselves as translators.In this process of learning and improving themselves as translators, they've come to realize that translation can be a joyful process and their confidence as translators can be boosted in their successful production of quality translation (Extracts 9 and10).
It can be seen from 6.2.1 and 6.2.2 that STIL, if welldesigned and implemented, constitutes a forced and facilitated process of subject learning and translation learning.On the one hand, the subject-focused part, while equipping students with subject knowledge, builds a cognitive and affective basis for subsequent translationfocused part.On the other hand, the translation-focused part, in cultivating translation ability, forces a critical closer engagement with ST and TTs, which enhances students' command and consequently admiration of the embedded subject knowledge.Thanks to STIL, the difficult and distant classical texts are made both accessible and relevant to learners and naturally their attitude toward them has changed.Such a change of attitude will probably inspire learners to approach other classical texts through translation activities, which can hopefully turn themselves over time into the specialists that translators are expected to be (Neubert, 2000).In short, the mutually supportive interaction between subject learning and translation learning, as discussed above, may recommend STIL as a more ready pedagogical innovation than its inspiration, CLIL, where the use of L2 might limit the width and depth of subject-learning and students' L1 expression of subject knowledge (Dalton-Puffer, 2007; Rose Mahan & Norheim, 2021).

Research Implications and Conclusion
While subject knowledge is almost universally recognized as part of translator competence, the systematic inquiry into its instruction in translation courses is quite scarce.In this study, STIL, a new approach to translation instruction characterized by integration between systematic subject instruction and translation instruction in a particular translation, is proposed, implemented, and explored.The results reveal learners' preference for such an approach for its positive role in both subject learning and translation learning.Such revelations carry specific implications for translation education.
Firstly, STIL may be a justified and even advisable approach to translation instruction.Emotionally, it is preferred by almost all participants to non-STIL approach experienced by them.Cognitively, there seems to be in STIL a mutually supportive interaction between subject learning and translation learning, with both contributing positively to rather than undermining each other.For the translation programs with a dual-focused aim of helping students acquire both subject knowledge and translation knowledge, STIL is evidently a preferable option.For those where gaining subject knowledge is not explicitly stressed, STIL is still worth trying for its contribution to translation learning.
Secondly, classics in a particular domain may advisably be chosen as the texts (both ST and TT) for implementation of STIL.Every particular domain has its own classics, which, together with its translation, can be of great pedagogical value in training translators (Reiss, 2014).However, there is a dearth of the application of classics to translation instruction, to which two main circumstance may be ascribed to: (1) supposed limited market demand for translation of discipline classics; (2) supposed high difficulty of classics for translation learners with no or limited domain knowledge.The first circumstance is generally a fact as commercial and technical texts do constitute the main body of current translation industry.However, translation is not only about business but also for communication of valuable knowledge created by human beings in different languages, which is even highlighted as the single corner stone of translation in Transknowletology, a new translation theory proposed by Yang (2021).Discipline classics are supposed to carry the most important knowledge in their respective domains, which shall and deserve to be communicated in quality TTs.Nevertheless, the poor translation of social classics, which is not rare at all, hinders the spread of such knowledge.Thus there is a potential huge demand for re-translation of discipline classics and consequently training focused on their translation.As for the challenge of classics as teaching materials for learner translators, this study suggests that the seemingly insurmountable difficulty of classics can be overcome in the process of STIL instruction, offering empirical support for using challenging tasks for translation instruction (Diamond & Shreve, 2017;Shreve, 2006).By extension, difficulty may not be an important consideration in selection of texts.More important than text difficulty might be text quality mainly measured by depth of subject knowledge and beauty of linguistic forms.Translators working on texts of this kind seem more likely to overcome the challenges met, experience happier translation process and produce better translation product.Thus, translation instructors shall, ideally, choose instructional materials not only of proper difficulties for them (Shreve, 2006), but also of subject and language to be appreciated or admired by them.Where it is impossible to have both good quality and proper difficulty, priority shall be given to quality.In short, teachers are encouraged to search in quality texts for those of proper difficulty.In this sense, classics are the right texts for translation education, especially for specialized translation education.
Thirdly, tasks shall be developed around the texts chosen to give a full play to pedagogical potential of classics.While it is true that translators will definitely gain some fragmented subject and translation knowledge by unguided translating itself, the proficient mastery of profound subject knowledge in a systematic way shall necessarily be facilitated by well-designed tasks.As for the development of tasks for STIL, it falls with translation teachers' proper business to design translation tasks but conceiving subject tasks is not traditionally their duty.This study suggests that the texts chosen shall be explored deeply to figure out subject learning tasks that enable learners not only to understand the basic ideas but also to establish their relevance to daily life and the domain development.All these efforts are made with the purpose of helping learners build up systematic and in-depth subject knowledge, which will render learners ready both emotionally and cognitively for subsequent translation tasks.This can be particularly important for translation courses where mastery of subject knowledge is prior to deep ST comprehension and quality TT production and can't possibly be well achieved in an ad hoc and frivolous way.Admittedly, the acquisition of subject knowledge adequate for development of such structured tasks is a great challenge for translation teachers with no or limited specialist background.This study shows that the challenge is not insurmountable and the efforts for overcoming it seem both worthy and necessary.By mastering the knowledge themselves, translation instructors can possibly attain or approximate the status of a domain expert and only by mastering it can they become capable of accomplishing effective integration between subject learning and translation learning, which is hard to achieve when a translator teacher, who is ignorant of the domain, is working together with a domain expert with little knowledge of translation.
Along with the positive results and implications are limitations of this study which shall be addressed in the future.Firstly, the implementation plan, though an outcome of continuous revisions over almost 5 years, is a contextualized one for a particular MTI course.It is advisable to explore other plans for different courses in different contexts in order to expand the knowledge base of optimal STIL practice.Secondly, the effect of STIL in this study is mainly based on learners' reports and more objective instruments can be used in future studies to further improve the reliability and validity of this line of inquiries.Finally, this study is focused on learners' perceptions of STIL and the probe into how teachers, the other major players in translation instruction, perceive STIL, especially its challenges, will be of equally great value for its application in translation education.
Before this course, I seldom doubted the translation quality of classical works.In this course, I found many problems in the published TTs, of which some are unbelievable.As a translation major, I long for correct and accurate translations that I can put myself into with complete joy and I've come to realize that translators perhaps matter more when it comes to translation of classical academic texts.As I examined the published translations and attempted to produce my own version, I perceived the artistry of the ST and became more confident in my translation ability.(TR,S13) Extract 10: I thought in the very beginning of this course that translating classical texts was beyond me.As the course carried on, I found the linguistic beauty and refinement of classical texts and at the same time, the charm and interest of their creative translations.It is worthy of the efforts and interesting to translate classical ST into creative TT that is concise and smooth in parallel structure and with rhymes.thetranslated version to see if there is any creativity, loss or errors in the translation.(SRI,S6) Extract 14: Creative translation consists in adopting ''novel'' ways to translate rather than being confined to the linguistic forms of ST.However, the precondition for creative translation is a certain level of subject knowledge.The idea of ST can only be represented in creative TT when we understand it deeply.Without such a comprehension, we would often find ourselves confined to the forms of ST and incapable of flexible transformation.(TR,S5) Extract 15: When I am familiar with or have a good understanding of particular chapter of WN, I am more likely to identify the errors and losses in translations.(TR,S6).Extract 16: I find my motivation for translation (learning) has been enhanced because of STIL.It has made me aware of my weaknesses in knowledge(cognition), logic and language, which escaped my attention before.(SCI,S4) comprehension.At the same time, I am made aware that I can approach the published TTs through the three translation tasks, which enhanced my liking for classical texts.For example, now when I am reading Principles of Economics by Mankiw, I firstly read the original text and then compare it with