Skip to content ↓

February 2020 - Year 22 - Issue 1

ISSN 1755-9715

Constructing Conversations: The Viability of Student-centered Pedagogy in Confucian Heritage EFL Classrooms

Ian Roth is a teacher at Meijo University.  He has written and co-written books and papers on the functioning and design of educational systems. His current professional interests include leadership dynamics and the interdependence of schooling and culture.  Ian enjoys working in all manner of educational settings.  Email:



This paper presents the details of and results produced by an attempt to bring student-centered pedagogy to EFL learners in Japan.  Research suggests, and based on personal experience I can confirm, that Confucian-heritage students often struggle with student-centered learning methods (Iyengar, 2010; Jackson, 2002).  They can be reticent to share their knowledge, opinions, and questions in classroom settings.  This paper describes the results of efforts to employ a student-centered pedagogy in such classrooms. During the course of the lessons here described, which spanned two class periods (180 minutes), the teacher spoke for approximately 10 to 15 minutes total. Rather than delivering content, the teacher acted as a facilitator, delineating the container in which the interactive process took place and providing the broad steps to guide it.  The ideas, content, and corrections were entirely provided by the students.

To simplify the identification of this process, it will heretofore be referred to as ‘constructing conversations.’  It asks the language learners to make heavy use of what they have already learned and is, therefore, best suited for classrooms that have some experience dealing with the second or foreign language being taught.  Having said that, this type of lesson does not require any particular level of skill—it is previous exposure rather than proficiency that allows the lesson to function.

The entirety of the constructing conversations cycle as it will be represented here required two ninety-minute class periods, though the first half could have been used as a standalone class.  The first period was dedicated 1) to deciding the context of the conversation and 2) to the construction of that conversation.  The second period was spent recreating and exploring the possibilities of the conversation.  The content of both lessons was almost entirely learner driven and derived.  The educator’s role was one of facilitation accomplished through the flexible application of a process and set of structures designed to predispose the learning group towards collaboration and the production of emergent linguistic and learning outcomes.


Theoretical basis of lesson design

There is more to the theoretical underpinnings of the process than a focus on learner-driven content.  It followed, in both iterative and nested ways, the learning steps that constitute a three-stage learning process I have called the Trans-Contextual Learning Model (Roth, 2019).  A meta-model extrapolated from models of learning found in diverse cultures, disciplines, and time periods (Ackoff, 1991; Hansen, 1989; Joseph, 2002; Kim, 2003; Pranin, 2005; Scharmer, 2009; Tan, 2015; Nietzsche, 2005), it identifies the learning stages of compiling, networking, and contextualizing.  Respectively and simply, these equate to 1) the accumulation of information pertinent to what is to be learned, 2) the ordering of that information into a coherent knowledge framework and stretching of that framework such that its limits are identified or expanded, and 3) the application and testing of that framework in a new, feedback-rich context such as that of teaching it or of solving a real-world problem. 

Though the stages are numbered for ease of identification, they can be engaged with as a flexible sequence rather than a normative series—each informs the others and movement between them should take place as needed.  Often during the contextualizing stage, the learner finds his or her knowledge or understanding lacking in some respect or to some degree.  The recourse is a return to one of the other stages.  In this way, the process can be circular and is often a spiral wherein repeated cycles build upon earlier ones.

The lessons employed in this study were based on these stages.  The learners began by identifying discrete elements that may have been of consequence in a given the situation and context.  At this point, these elements had not been fitted into a larger, coherent scheme.  The creation of such a scheme took place during the next stage.  Redundancies were eliminated, gaps were filled, and alternatives were identified.  In the third stage, this scheme was pressure tested through progressively more demanding situations in which the supports were removed and confounding elements were introduced. 


The lessons

The first class


The constructing conversations process begins with a question.  Though it is not the only suitable one, the data contained herein was produced by the question, ‘What are some situations in which you might have to use English?’  As the italics suggest, the you element of this question was stressed—the focus was on the learners themselves and the reality of their lived experiences.  The goals of posing this question were for the learners to determine the contextual focus of the lesson and for the activity to address realistic concerns they may have had with respect to using the language being learned.  A prerequisite of achieving these goals was that the context being addressed was relevant to the learners who actually populated the classes rather than to the hypothetical challenges faced by an imagined individual. 

The question of when the students might have to use English produced several context-driven situations.  The learners then collectively determined which one they found most interesting and/or anxiety-producing.  This determination was made based on a vote.  Example contexts and situations that groups have chosen in the past include losing one’s wallet, checking into hotels, navigating airports in foreign countries, helping with directions, and acting as a tour guide in one’s own country.  When the context did not by nature make it obvious, the role that the learners saw themselves occupying was also determined (e.g. in the case of a restaurant, the learners envisioned themselves as customers rather than servers). 



Once the context had been determined, the groups were asked to write on the board a specified number of phrases they might use in the that context.  The required numbers were facilitator-determined and took the contexts and size of the classes into account.  For each, it was between fifteen and twenty—high enough to make it appropriately challenging as a goal, but not be so high as to make the resulting conversation unwieldy.  A further stipulation was that the phrases had to all be drawn from the learners’ side of the conversation.  In the context of ordering in a restaurant, this equated to phrases such as ‘what is today’s special’ being admissible while ‘are you ready to order’ was not.


Correcting (Optional)

At this point, depending on the level of the students, the goals of the lesson/course, and the amount of time remaining, time may have been spent correcting errors in the produced phrases.  As with the rest of the lesson, this was accomplished with as little input from the facilitator as possible by either 1) giving the classes the total number of errors present, or 2) identifying sentences that contained errors.  In either case, the learners were tasked with making the final error identifications and corrections.



The next stage consisted of the learners putting the phrases they produced in logical order based upon when they would likely be used.  For example, ‘Non-smoking please’ was placed close to the beginning, ‘Check please’ near the end, and ‘I’d like a coffee, please’ somewhere in the middle.  The completion of this ordering task was left entirely to the learners.  Each time it produced a different classroom dynamic (e.g. one student writing and the others advising, two students guiding while many wrote, etc.) and a sensible, but largely unpredictable ordering of the produced phrases.  During this stage, the groups were also tasked with adding the phrases needed to resolve any apparent gaps in the conversations. 



During the next phase of the lesson, the learners were asked to fill in the other side of the conversation (e.g. the waiter or waitress that they as customers would be dealing with).  In that they had to consider which pieces of conversation would naturally follow and lead into the existing pieces as they were ordered, this stressed the same skillset as the gap-bridging just described.  They were again challenged to identify and correct any grammatical or spelling errors.  Once completed, the group had an entirely learner-created, two-part conversation.  They practiced this conversation with whatever time remained in the period.


The second class


The second class started with the learners attempting to reconstruct the conversation from the previous class without the use of any notes or reference materials.  After a few minutes of this, they were permitted to use their notes and expected to reconstruct the conversation perfectly.  Once the complete conversation was available to them, the learners went through it with a partner, playing each role once.  The partners then quizzed each other on the content by taking a line from the conversation and employing it as a prompt for the previous or following line.  After a few minutes of quizzing, the pairs attempted the conversation again, relying mainly on their memories and referencing the board as little as possible.



The goal of the next stage was to introduce flexibility into the conversations.  The groups’ task was to add several possible variations to the conversations.  These variations could be minor, such as changing the number of people in one’s party, or could amount to entirely new branches of the conversation.  Once sufficient variation had been introduced and any errors in the new elements corrected, the learners repeated the sub-process that closed the prior stage—practicing the conversation, quizzing the content (this time providing multiple possible responses), and trying to perform the conversation from memory.  This stage required increased responsiveness to one’s partner as, depending on the variations that person employed, the script may have been a misleading guide.



During the final phase of this activity, the facilitator determined a time that was slightly longer than it took the learners to complete the conversation when working from memory.  He then erased the last third of the conversation.  The learners were tasked with speaking continuously until the time ran out.  They could follow what they remembered of the erased script, but generally needed to improvise a bit.  The middle third of the script was erased next and the process repeated.  Then, everything except for the first line of the script was erased and the process was performed again.  In the final set of evolutions, the facilitator introduced a word, or multiple words on subsequent iterations, that had not been present in the script.  The learners were tasked with incorporating the word(s) into their conversation.  At this point, they were employing a structured but improvisational approach to the situation—aware of their goals and of what was needed to achieve them, but forced to find unplanned routes to the places from which those goals could be achieved.


Data collection and results

I conducted the above lesson set with four different classes of university students numbering 19, 22, 15, and 16 (n=72).  The ability levels of the classes varied.  The situations they chose to address were ‘arriving at a foreign airport’, ‘in a foreign restaurant’, ‘on a train’, and ‘giving direction’.  I used anonymous sentence stems pre- and post-activity to elicit feelings about the prospect of having to use English in each class’s chosen situation.  Though even a cursory before-and-after comparison of the learners’ language use would make apparent their having improved, I was primarily interested in the degrees to which the learner’s would become comfortable with the prospect of facing that situation and with the mindset changes these might suggest.    

The stem I used to elicit the learners’ attitudes about the situation was, “If I had to use English in (the chosen context/situation), I would feel…”.  In completing this stem, they were encouraged to answer with one or more feelings.  Prior to the constructing conversations exercises the most common responses were:



Times written:











Given the negative affect associated with four of these responses, the number of times ‘excited’ was written is bit surprising.  When all responses were included and coded as examples of positive, negative, or neutral (cannot be determined) valence, the pre-activity response numbers were as follow:

Positive valence


Neutral valence/Cannot be determined


Negative valence


The positive valence completions were entirely accounted for by the ‘excited’ responses (-11) and one ‘happy’ response.  Among others the negative valence responses included ‘afraid,’ ‘panicked,’ ‘embarrassed,’ and ‘confused’.  The neutral valence/cannot be determined responses were ‘pounding heart’ (-2) and ‘surprised’ (-4).

These responses suggest that the prospect of dealing with the situation in question was generally perceived by the learners to be unpleasant if not threatening.  The overall impression to which these responses contribute is that an encounter, such as the ones chosen, that requires the use of English is something to be avoided if possible.

Following the activity, the learners were asked to complete the same sentence stem for a second time.  This time, the most common responses were:


Times written:











Despite the fact that nervous remained among the most common responses, its post-lesson prevalence was less than one-fifth that of its pre-lesson prevalence.   Even more drastic are the changes in valence.

Positive Valence


Neutral Valence/Cannot Be Determined


Negative Valence


The ratio of positive to negative in the pre-lesson responses was 0.16.  For the post-lesson responses, it was approximately 4.43.  Following are three of the conversations that the classes were able to construct by the end of the first class with only the most limited teacher input:

Context- Restaurant


A: How many?

B: We have three people.

A: Smoking or non-smoking?

B: Non-smoking.

A: Please sit in these seats.

B: May I have a menu please?

A: Sure.  Here you are.

B: Please tell me what’s for lunch.

A: We have pizza and hamburgers.

B: What do you recommend?

A: Everything is very good.

B: What is the most popular food?

A: The pizza is the most popular.

B: Is it spicy?

A: It’s not spicy.

B: What is on the pizza?

A: Ham and tomatoes.          

B: I want to order.

A: What would you like?

B: I want to eat pizza please.

A: What would you like to drink?

B: (Drink) please.

A: Is everything okay?

B: Where is the bathroom?

A: It’s over there.  Can I get you anything?

B: I would like water.

A: Here you are.

B: How much is it?

A: It’s free!    

B: Check, please.

A: Thank you.

B: Thank you.


Context- Arrival at an airport in a foreign country


A: We will arrive in 10 minutes.

B: It’s about time to arrive.  I’ll enjoy traveling.

A: Have a nice day!

B: Where is the exit?

A: There it is.             

B: Thanks.

C: Please show me your passport.

B: Here’s my passport.

C: What is the purpose of your visit?                       

B: My purpose is sightseeing.

C: Do you have any dangerous things?        

B: I don’t have any dangerous things.

C: How long will you stay?              

B: I will stay one week.

C: How many times have you visited (country)?      

B: It’s my first time.

C: Where are you from?

B: I’m from Japan.  Where is my baggage?

C: Show me your ticket.  Wait at number 7.  Have a nice trip.        

B: Thank you.  Where is the rest room?

C: It’s over there.     

B: Where can I change money?

C: You can change at an exchange counter.

B: Do you know any famous shops around here?

C: No, sorry.  I don’t.

B: Where is a tax-free shop?

C: Go straight along this street.                       

B: Where is the bus for my hotel?

C: It’s there but the next bus will come in one hour.    

B: Where can I get a taxi?

C: Go straight on this road.



Context: On the train


A: Excuse me?

B: Yes?

A: Which train should I get on?

B: Where do you want to go?

A: I want to go to Shibuya.

B: You have to take this train.

A: Where is this train bound for?

B: This train is bound for _____.

A: Which station should I get off?

B: Please get off the train at _____ station and get on the train bound for _____.

A: How many stops is it?

B: It’s the third stop.

A: Tell me more.

B: Transfer to the _____ line.

A: What time next train arrive at here?

B: Next train arrive at here after 5 minutes.

A: How long does it take?

B: It takes about _____ minutes.

A: I can’t understand.

B: I take you.

A: Thank you.  Anyway, let’s eat lunch.

B: Sorry, I don’t have enough time.

A: Okay.  Do you know good restaurant near here?

B: Sorry, I don’t know.

A: Really?

B: I’m a stranger here.

The third of these conversations contains grammatical errors as there was insufficient class time to have the students correct them.  During the second class, the students added options to each of these conversations, making them into branching paths.  Here is an example of what this looks like from the fourth class.

Context: Giving directions


A: I need your help.

B: What are you looking for?

A: I would like to go to (_____ University/Nagoya Castle).

A: I’m a (stranger here/_____ student).

A: But I need your help!

B: I’m not sure about _____, but I’ll check.

A: Thank you!

B: How do you want to get there?

A: Can I (see a map/ask a station attendant)?

B: (You are here/Leave it to me).

A: Okay.

B: I recommend that you take the (train/bus).

A: Which train should I take?

B: Please take the _____ line train.

A: How much does it cost?

B: It’s _____ yen.

A: Oh no! (I have no money/I want to save money).

B: (You had better walk/I’ll pay for you).

A: Sorry?

B: You can walk there.  Let’s go together.

A: Really? Thanks!! How long does it take?

B: It takes _____ to get there.

A: Where is it located?

B: It is next to ____.

A: Which way should we go?

B: Go straight this way.

B: Okay.

B: Please follow me.



These constructed conversations can be characterized in several noteworthy ways.  None of them are particularly clean.  Rather, they are a bit messy, containing several disjointed sections and odd inclusions.  As opposed to the pre-composed conversations that are typically found in EFL/ESL textbooks, the above conversations are patchwork-like because they represent the efforts of a group of learners to collectively map what was previously, for each individual, largely unfamiliar territory.  They also contain moments of humor such as the very end of the third conversation when the person giving directions admits to being a stranger in the area.  Such instances are indications that the learners were engaged in language play.  Given what began as an incoherent collection of phrases, they had to imagine multiple possible scenarios, test them out, and determine which provided the most sensible, accommodating, and appealing through-line.  This was done repeatedly as each individual learner’s choice of where to place a given phrase compelled the other learners to rework their imagined scenarios.

The playful character of this process must be one element that produced the changes evidenced by the pre- and post-activity sentence stem responses.  Every student represented here had extensive experience learning pre-determined conversational patterns.  This is precisely the knowledge base on which the constructing conversations process relied when eliciting contextually appropriate phrases from the learners.  But, beyond that point, the process was one of wrestling with the language and its in-context possibilities, of advancing, retreating, and correcting, and of negotiating and renegotiating with other learners.  One result of this process appears to be greater positivity and poise when considering the uncertainty and complexity of real-world, second-language interactions.  

In that the final conversation is a product of an active, democratic process, the learners were exposed to a much richer data stream than they would have been through traditional teaching methods.  Their individual ideas coalesced into a collective determination; their individual knowledge interacted to produce collective knowledge.  By working through these interactions, the learners were collaboratively working for the knowledge that resulted.  In this way, the learners’ relationships with the content increased in complexity and the learning was personalized (Palmer, 1993).      



The conversations produced by these classes and the shift in their sentence stem responses pre- and post-activity suggest that this process produced positive outcomes.  I am unable to provide a set of analogous data produced by a more traditional lesson but, even lacking this, what has been presented here suggests that student-centered approaches can be functional even in contexts where they are abnormal.  The process detailed here relied almost entirely on the learners, leveraging the concealed knowledge of the group and transforming it into revealed knowledge put-into-practice by the individuals.  The role of the teacher was that of a facilitator, who acted only to create and modify the container within which the collaborative process played out.  The findings here presented suggest that activities such as this may be viable means of engaging even (culturally) reticent learners in a student-centered learning process.



Ackoff, R. (1991). Ackoff’s Fables: Irreverent Reflections on Business and Bureaucracy.

New York: Wiley.

Hansen, C. (1989). Language in the heart-mind. In Robert E. Allison (ed.) Understanding the

Chinese mind: The philosophical roots. Hong Kong: Oxford University Press, pp. 75-124.

Iyengar, S. (2010) The art of choosing. New York, NY: Hachette Book Group.

Jackson, J. (2002). Reticence in second language case discussions: Anxiety and aspirations. System, 30(1), 65–84.

Joseph, M. (2002). The trivium: The liberal arts of logic, grammar, and rhetoric.            Philadelphia,

PA: Paul Dry Books, Inc.

Kim, H. K. February, (2003). Critical thinking, learning and Confucius: A positive assessment.

In Journal of philosophy of education, 37 (1), pp. 71-87.

Nietzsche, F. (2005). Thus spoke Zarathustra. New York, NY: Barnes & Noble Classics.

Palmer, P. J. (1993). To know as we are known: Education as a spiritual journey. San Francisco,

CA: Harper Collins.

Pranin, S. (2005). An interview with Endō Seishirō Shihan. Nishina, D. & Hideo, A. [Trans.] In Aiki News, Japan No. 144

Roth, I. (June, 2019). An attempt at a trans-contextual learning model: Extracting a meta-modelacross time, space, and discipline. Integral Leadership Review.

Scharmer, C. O. (2009). Theory U. San Francisco, CA: Berrett-Koehler Publishers. Tan, C. (2015). Beyond rote-memorisation: Confucius’ concept of thinking. In Educational Philosophy and Theory, Vol. 47(5), pp. 428-439.


Please check the  Creative Ways to Get Students Speaking More course at Pilgrims website.

  • Does Conversational Interaction Really Matter?
    Ali Maskari, United Arab Emirates

  • Constructing Conversations: The Viability of Student-centered Pedagogy in Confucian Heritage EFL Classrooms
    Ian Roth, Japan