An Independent Variable and a Dependent One

[kellogg]

What evidence is there that intra-mental knowledge emerges from the use of inter-mental skills in this data?

Ⅰ. Introduction

In this data, I will search for evidence of intra-knowledge emerging from intra mental skills examining utterances. Both of the data are from regular English class taught by an expert English teacher to fifth graders. The discourse shown in data 1 is casual greeting talk for initiating a new lesson. On the other hand, data 2 shows practise of a fixed expression‎, merely varying in vocabulary level. Since both the data are strongly teacher led type, I will examine teacher's utterances to clarify the characteristics. I will compare the teacher's utterances shown in the two data operationalizing them by reversible roles. I will use <Teaching Talk as a Game of Catch - by Kwon and Kellogg> (Use quotation marks!) as a theoretical base, and sort out the utterances as reversible, and non-reversible roles. And I will subdivide them into projective commands, display questions and set texts as non-reversible roles and greetings, referential questions, offers and co-constructed texts as reversible roles.

Utterances can be examined for many things.

We've seen that Jisu examines them for concepts. Minkyeong examines them for concrete reference. Seong-eun examines them for intersubjectivity, for a theory of mind, Eunshil examines them for grammatical complexity, and so on.

What is Yura looking for? Let's see: she's looking for nonreversible roles. She's looking for commands, display questions and set texts. She's also looking for reversible roles: referential questions, offers, and co-constructed texts.

That's a LOT. But Yura can handle it, because she's sorted the six constructs into two basic categories: reversible and nonreversible. The problem is that "reversible" and "nonreversible" are inter-mental categories; they are ways of looking at interaction. So how will she show intra-mental knowledge emerging?

Remember that you need TWO variables: an independent, predictor variable (in this case, skill) and a dependent, predicted outcome variable (that is, knowledge in this case). The whole problem is to show how one type of variable is related to the other, how knowledge emerges from skill, how intramental understanding arises in inter-mental interaction.

If you only have ONE variable (utterances of different types) you will end up with an answer that is phenomenological, that is, DESCRIPTIVE, but not PREDICTIVE. This will give you INTERNAL validity, but it will not give you EXTERNAL validity.

Ⅱ. Theoretical Background

I will be examining the data by asking if the teacher's role is reversible to the students. Reversibility of roles can be defined as 'possibility of being reused by the other party' or 'availability of roles to the other party'. If we refer to an extract in 'Kwon and Kellogg', in the ESL specialist's class, the roles of information giver and information getter were not as reversible as those in the general subject teacher's class. With 'Uptakes' and 'Syntagmatic Variation', the reversibility of roles is one of the factors which affect whether the exchange is predictable.

Yura is starting to solve her problem. She has to show that in some way her independent variable (reversibility) is related to a dependent variable (knowledge). Right now there's a gap.

She introduces another variable (predictability). Let's see if this works to bridge the gap. If she can answer these TWO questions "yes", then she's done her job.

a) Is predictability CLEARLY related to the independent variable?

b) Is predictability CLEARLY related to the dependent variable?

The answer to a) is clearly YES, for the reasons that Yura gave above. If something is reversible, then we can predict that it will (eventually) be reversed. And if it is reversed, we can predict pretty accurately what the new reversed utterance will be. So predictability is related to reversibility.

But the answer to b) is not so clear. We have to somehow show that predictability is related to KNOWLEDGE. Hmmmm.... To UNDERSTANDING.

There can be various reasons of the roles being reversible/non-reversible, but I will approach by classifying the reasons into functional aspect and degree of difficulty.

How does this relate to all the different categories Yura laid out for us earlier? Remember?

"What is Yura looking for? Let's see: she's looking for nonreversible roles. She's looking for commands, display questions and set texts. She's also looking for reversible roles: referential questions, offers, and co-constructed texts."

All of these look like functional categories to me. Right?

In a functional approach, to refer to 'Kwon and Kellogg', roles can be reversible when they are greetings, referential questions, and offers. To take some examples from my data,

KT: Hello, everybody! (Greeting, buy professor! I have a question. Once in our class, we discussed if saying hello at the beginning of the class is really reversible. We made our conclusion that it's NOT, because its purpose is to initiate the class.(Should be the teacher's role) So in this case, it's not reversible, right?)

Right. But then it's not a greeting. It's an initiation. Right?

Ss: Hi/ Hello. Miss Sadie.

KT: How are you today? (This is reversible though?)

Can the children say it? Can they say it first? (Not: DO they say it first. But COULD they say it first?)

Ss: Good, Soso, Not bad..@*%#

S1: I'm bad.

KT: Oh, Why are you feeling bad today?

S1: 그냥..bad condition.

...........

KT: You have English Class, and..? (co-construction)

S4: Math.

............

KT: How about this picture? Who would like to make a sentence? Guen-hee? Guen-hee: (She was not paying attention) .........

KT: Can you make a sentence? (Offering)

Why is this an "offering"? If this is an "offering", then isn't the previous initiate, "Who would like to make a sentence?" also an offering?

Guen-hee: Girl. Pretty.

KT: What a? (co-construction)

Guen-hee: What a girl pretty? (Her partner corrects her) What a pretty girl!

This actually makes it look like CORRECTIONS are reversible! Amazing!

Guk and Kellogg found that corrections were reversible in the sense that children could use them TO OTHER CHILDREN. But they were NOT reversible in the sense that the children could NOT use them to the teacher.

Perhaps we need THREE categories? Fully reversible, Fully nonreversible, and reversible to other kids but not to the teacher.

On the other hand, when the roles are projective commands, display questions and set texts, they are non-reversible.

Good.

For example,

KT: Okay. Now open your book to page 20. (projective commands)

Ss: (open their books)

KT: Let's read. There are words from Sunday to Monday. I want you to read after the CD first. (projective commands)

Ss: (reading aloud)

..............

KT: What a green tree, but see, Jae-hyuk! Is there only one tree in this picture? Or are there many trees? (display question)

Ss: Many Trees.

The reversibility of roles can be determined by their degree of vocabulary/ grammatical difficulty and their complexity.

The word "role" is a little confusing here. "Role" is functional.　Remember that we used "reversible utterance" for lexicogrammatical reversiblity.

If an utterance is behind the students' grammatical or vocabulary level, or the complexity is too high in length, it will lose its reversibility. To take each one of examples which is reversible and non-reversible from my data,

FT: I'm not so well. I have a bad cold. And my throat really hurts.

->Reversible with easy grammar and relatively short meaning length.

...................

KT: Alright. Next one. (The next picture is of a blue sky above a peaceful field.)

Well, this one seems to be a little difficult to make a sentence, but there's no right answer.

->non-reversible due to using difficult grammar like 'seems to be', and also because the meaning length is comparatively long.

Good. But...let's go back to the issues we raised at the beginning of this section. Remember? We asked how "reversibility" and "predictability" were related to our independent and dependent variables. We could see that predictability is linked to the independent variable, that is, to inter-mental skill. But what about knowledge? Have we established a link?

Ⅲ. Hypothesis and Results

In this study, I will analyze 2 pieces of data by dividing them into 4 units each. Data 1 was split into 4 sequences in term of the topics, and data 2 was split into the same number of sequences but by whether the discourse is of talk or meta talk.

Notice that Yura changes tense in a rather disconcerting way. Her FIRST sentence is in the FUTURE tense. But her SECOND sentence is in the PAST tense. Let's see what the next one says....

I examined the teachers' utterances on the basis of Kwon and Kellogg, but with a revised classification. The classification below is what I adopted for analyzing my data. Also, I used some statistics to see the number and proportion of each classified utterances in each sequence and data.

RR: reversible role (greetings, ordinary talks, offers, co-constructions)

NR: non-reversible role (beginning the class, projective commands, nominations, display questions, rebroadcastings, eval‎uations, requiring students' reaction, giving new language information, explanation about the task)

RU: reversible utterance (mean length shorter than 7 words, grammatically or lexically reversible)

Good! But then shouldn't you use this above, where you say "reversible role" for lexicogrammatical reversibility?

NU: non-reversible utterance (mean length longer than 7 words, grammatically or lexically beyond the students' level)

Data 1

Date: April 3rd, 2009

Students: 5th grade, 26

Lesson: 2. What Day Is It Today?

Teacher: Shin Yu-ra(Korean teacher), Sun Lee(Foreigner co-teacher)

sequence 1-----------------------------------------------------

KT: Hello, everybody! NR(beginning the class) RU

Ss: Hi/ Hello. Miss Sadie.

KT: How are you today? RR(greeting) RU

Ss: Good, So so, Not bad..@*%#

S1: I'm bad.

KT: Oh, Why are you feeling bad today? RR(ordinary talk) NU

S1: 그냥..bad condition.

KT: Oh, You are not feeling very well. NR(rebroadcasting) RU

S1: Yes.

KT: How are you, Mrs. Sun? RR(greeting) RU

FT: I'm not so well. RR(ordinary talk) RU I have a bad cold. RR(ordinary talk) RU And my throat really hurts. RR(ordinary talk) RU

KT: Oh, that's too bad. RR(ordinary talk) RU

S2: That's too bad.

sequence 2-----------------------------------------------------

KT: Class, what day is it today? NR(requiring students' reaction) RU

Ss: (pause) .. It's Friday.

KT: Yes. NR(eval‎uation) RU It's Friday today. NR(rebroadcasting) RU Do you like Fridays? RR(ordinary talk) RU

Ss: yes/ no.

KT: Ji-seung, why don't you like Fridays? RR(ordinary talk) RU

S3: Because, many academies.

KT: You have to go to many academies? NR(rebroadcasting) NU How about you, Hyung-bo? RR(ordinary talk) RU

S2: I don't like it too.

KT: Why? RR(ordinary talk) RU

S2: Because I have 6 periods.

sequence 3-----------------------------------------------------

KT: You have 6 periods. NR(rebroadcasting) RU Long day! RR(ordinary talk) RU What classes do you have today? NR(requiring students' reaction) RU

Ss: ..........

KT: You have English Class, and..? RR(co-construction) RU

S4: Math.

S5: Korean. We have Korean everyday.

KT: You have Korean everyday? NR(rebroadcasting) RU

S6: Music!

S1: PE! 2교시 is P.E.

KT: Wow, you have P.E today! NR(rebroadcasting) RU Does everybody like P.E? RR(ordinary talk) RU

Ss: Yes.

KT: So you have English, P.E, Math, Korean, Music and..RR(co-construction) NU what else? RR(ordinary talk) RU You have one more class. RR(ordinary talk) RU

Ss: 도덕.

S4: 도덕이 영어로 뭐지?

KT: Moral Education. NR(giving new language information) NU You have Moral education today. NR(giving new language information) NU

Ss: (Some students) Moral education.

FT: Or Ethics. NR(giving new language information) NU

sequence 4-----------------------------------------------------

KT: Okay. Now open your book to page 20. NR(projective commands) NU

Ss: (open their books)

KT: Let's read. NR(projective commands) RU There are words from Sunday to Monday. NR(explanation about the task) NU I want you to read after the CD first. NR(projective commands) NU

Ss: (reading aloud)

KT: On the bottom, there's a word snail. NR(explanation about the task) NU And Sunday, Monday, Tuesday.. Seven days are hidden in the snail. NR(explanation about the task) NU Let's find the words and circle them. NR(projective commands) NU Is everybody ready? NR(requiring students' reaction) RU

Ss: Yes. @*#& 쉽다.

S1: 벌써 3개 찾았어요.

-----------------------------------------------------

	RR	NR	RU	NU	total number of utterances
sequence 1	7 (77.8%)	2 (22.2%)	8 (88.9%)	1 (11.1%)	9 (100%)
sequence 2	4 (50.0%)	4 (50.0%)	7 (87.5%)	1 (12.5%)	8 (100%)
sequence 3	6 (46.2%)	7 (53.8%)	9 (69.2%)	4 (30.8%)	13 (100%)
sequence 4	0 (0%)	8 (100%)	2 (25.0%)	6 (75.0%)	8 (100%)

In data 1, the first sequence is the usual greeting at the very beginning of the class. Since the exchange is casual and consists of everyday talk, a large portion of teacher's utterances are reversible. The topic of sequence 2 is about days(Sun-Sat) while sequence 3 is about subjects, so we can see that the teacher is intending to lead the class to the topic of the lesson.( 2. What Day Is It Today?) The teacher seems to use a lot of reversible utterances in terms of grammatical, lexical or meaning length, while the role is not always reversible as the teacher's role here is to ask set of questions to lead the students to the main topic. At sequence 4, none of the roles are reversible and merely quarter of the utterances are reversible. This is because the teacher is getting the students do their task on their textbooks individually, so a fair part of the discourse is of teacher's commands or explanation about the task.

Good. But can we link that to knowledge?

In the data, we can find some evidence of intra-mental knowledge emerging from inter-mental skills. In sequence 1, we can see S1's knowledge by use of the word 'condition' which seems to come from Korean idiomatic expression‎.(It can be regarded as Konglish). But it can be evidence of integration of the native language and the foreign one, also integration of old vocabulary with a new grammar.

Right now, Yura has some trouble. Her data is FAR AWAY. This makes it hard for her to do this kind of content-based analysis, and it makes it even HARDER for the reader to check her work.

Let's reorganize this a little. Right now, Yura has the data AND the coding. Then she presents tables based on the coding. Then she discusses them.

What happens if we do this?

a) Tables.

b) Discussion of the tables.

c) Raw data (with or without codings)

d) Discussion of the data.

It seems to me that if you reorganize it this way you will have the NUMBERS closer to the quantitative analysis and the RAW DATA to the qualitative analysis.

In the third sequence, we also can find out the vocabulary level knowledge is emerging as they name the subjects. Besides, we can find an sounding level knowledge when the students repeated the word 'moral education'. This is also integration of sounding and meaning in the aspect of learning a new vocabulary connected with their native language(도덕).

Try using QUOTATIONS from the data (using quotation marks and parentheses).

From the analysis, we can see that the reversibility has declined in sequence four. Which kind of reversibility? Functional or lexicogrammatical? This suggests that the children are working on internalization rather than externalization in sequence four.

Really? Does it always suggest that? When the teacher calls the attendance roll, the roles are not reversible. Does that mean the kids are internalizing?

Perhaps "suggest" is the wrong word here. Perhaps "suggest" is a strong word here. It's not the same thing as "might mean".

However, this doesn't prove that more intra-mental knowledge was emerging in sequence 4 even if internalization was prominent. It doesn't? Isn't that exactly what internalization means? Rather, we could find more evidence of intra-mental knowledge emerging from inter-mental skills in sequence 1-3. What exactly is the evidence? Is the reversibility of utterances evidence of knowledge AND evidence of skill? How?

You can see that Yura never quite solved the problem of connecting predictability to UNDERSTANDING. Instead, we have a huge number of categories, descriptions, and classifications...but not EXPLANATIONS.

Try it like this:

a) My independent variable is reversiblity (in both forms: utterance reversibility and role reversibility)

b) My dependent variable is PREDICTABILITY (that is, uptake and understanding).

c) Reversibility is evidence of inter-mental skill, because listening becomes speaking.

d) Predictabiity is evidence of intra-mental knowledge, because listening becomes understanding.

Go through the data exchange by exchange, looking for reversiblity and then understanding.

Now, you can see that Yura jumps to the next bit of data. This is a bit of a leap; it's not clear WHY we are doing this. Will it give us greater EXTERNAL validity? If so, how?

Data 2

Date: April 23rd, 2009

Students: 5 grade, 27 students

Lesson: 4. What a Nice Day!

Teacher: Shin Yu-ra(Korean teacher)

sequence 1-----------------------------------------------------

KT: Alright, let's make sentences looking at these pictures. NR(projective commands) NU You have to make sentences describing the picture, using 'What a -'. NR(explanation about the task) NU Okay? NR(requiring students' reaction) RU

Ss: Okay.

(Showing the first picture with ppt, )

sequence 2-----------------------------------------------------

KT: who can make a sentence about this picture? NR(requiring students' reaction) NU

(Students goes mumbling about the picture.)

KT: Seung-ho? NR(nomination) RU

Seung-ho: What a tall man!

KT: What a tall man! NR(rebroadcasting) RU Good! NR(eval‎uation) RU I'll write it down. RR(offer) RU (typing the sentence that Seung-ho's just made on the PPT slide). Everybody, looking at the picture, let's read the sentence altogether. NR(projective command) NU

Ss: What a tall man!

KT: How about this picture? NR(requiring students' reaction) RU Who would like to make a sentence? NR(requiring students' reaction) NU Guen-hee? NR(nomination) RU

Guen-hee: (She was not paying attention) .........

KT: Can you make a sentence? NR(requiring students' reaction) RU

Guen-hee: Girl. Pretty.

KT: What a? RR(co-construction) RU

Guen-hee: What a girl pretty? (Her partner corrects her) What a pretty girl!

KT: Good job, Guen-hee. NR(eval‎uations) RU (Types in the sentence under the picture)

Everybody, let's read it together. NR(projective command) RU

Ss: What a pretty girl!

KT: Alright. NR(eval‎uation) RU Next one. NR(projective command) RU (The next picture is of a blue sky above a peaceful field.) Well, this one seems to be a little difficult to make a sentence, but there's no right answer. NR(explanation about the task) NU You can express what you think about the picture. NR(explanation about the task) NU Who would like to volunteer? NR(requiring students' reaction) NU Jae-hyuk? NR(nomination) RU

Jae-hyuk: What a green tree!

sequence 3-----------------------------------------------------

KT: What a green tree, NR(rebroadcasting) RU but see, Jae-hyuk! NR(projective command) RU Is there only one tree in this picture? NR(display question) RU Or are there many trees? NR(display question) RU

Ss: Many Trees.

KT: In that case, you shouldn't use 'a' because there's not only one tree. NR(giving new language information) NU Then what should we say? NR(requiring students' reaction) RU 'What..?'RR(co-constructions) RU

Ss: ...?

Jung-hyun: green trees!

KT: Excellent! NR(eval‎uation) RU Just like that! NR(eval‎uation) RU (Typing 'What green trees!')

Ss: What green trees!

sequence 4-----------------------------------------------------

KT: Can anybody make a different sentence about this picture? NR(requiring students' reaction) NU

Ss: .......

KT: Well, I see a very beautiful blue sky here, and the weather seems very nice.. NR(explanation about the task) NU

Yeah, Su-bin? NR(nomination) RU

Su-bin: What a beautiful sky!

KT: Very good. NR(eval‎uation) RU (Typing the sentence)

Everybody together! NR(projective command) RU

Ss: What a beautiful sky!

-----------------------------------------------------

	RR	NR	RU	NU	total number of utterances
sequence 1	0 (0%)	3 (100%)	1 (33.3%)	2 (66.7%)	3 (100%)
sequence 2	1 (5.0%)	19 (95.0%)	13 (65.0%)	7 (35.0%)	20 (100%)
sequence 3	1 (11.1%)	8 (88.9%)	8 (88.9%)	1 (11.1%)	9 (100%)
sequence 4	0 (0%)	5 (100%)	3 (60.0%)	2 (40.0%)	5 (100%)

In data 2, sequence 1 is an exchange of meta-talks, so to speak, are explanations or information about what language the students are supposed to produce. The teacher's utterances shown in sequence 3 are also meta-talk explaining grammatical knowledge. sequence 2 and 4 are exchanges of students' activity making sentences in a set of grammar, (item-based combinations)

It would be good if Yura could put in a REFERENCE here.

Otherwise it looks like:

a) Yura is claiming that SHE invented the category of item-based combination

b) Yura thinks EVERYBODY knows what this means, so she doesn't have to define it.

Remember that Yura BEGAN this answer with a very different set of concerns: reversibility (functional and lexicogrammatical), predictabiilty (uptake and understanding), and so on.

Is she completely changing her focus? Or is this just a quick reference? If she's SERIOUS about using this as a dependent variable, then she's going to have to go back and rewrite her theoretical background section, using Tomasello.

and the teacher's role is mainly commanding, requiring students' reactions or eval‎uation. Thus the teacher's role is hardly reversible even though pretty much part in sequence 2 were reversible utterances , due to using little grammar.

Intra-mental knowledge that are emerging from inter-mental skills, in data 2, is mainly in wording level. Because the activity in this data is building sentences using a fixed formula, the students are using their knowledge about grammar.

For example,

KT: who can make a sentence about this picture?

(Students goes mumbling about the picture.)

KT: Seung-ho?

Seung-ho: What a tall man!

KT: What a tall man! Good! I'll write it down. (typing the sentence that Seung-ho's just made on the PPT slide). Everybody, looking at the picture, let's read the sentence altogether.

Ss: What a tall man!

knowledge in the level of meaning doesn't show in this data because it's a grammar practice rather than building a story.

Remember that we saw TWO different kinds of meaning: semantic and pragmatic. Are stories necessary for BOTH of them?

From the analysis of data 2, we can observe a tendency that when the teacher's utterance is META-TALK, it is hardly reversible while it's the opposite in the other case.

You mean that it's ALWAYS reversible when it is not meta-talk? But if tha is true, then there is no such thing as lexicogrammatical nonreversibility. Reversibility is ONLY functional. Now, I‘m a little lost.

I would (tense) compare the two pieces of data by comparing the number of RR, NR, RU and NU and the portion in total number of teacher's utterances. We can see that the roles were much more reversible in data 1 than in data 2, while the reversibility of utterances were pretty much the same in the two.

	RR	NR	RU	NU	total number of utterances
DATA 1	17 (44.7%)	21 (55.3%)	26 (68.4%)	12 (31.6%)	38 (100%)
DATA 2	3 (8.3%)	33 (91.7%)	25 (69.4%)	11 (30.6%)	36 (100%)

The table below is comparison of numbers of roles and utterances which are reversible, and either the role or the utterance is non-reversible. From the statistics, we can see that the reversibility of roles and utterances are about 5 times more in data 1(which consists of greeting and everyday talks) than in data 2(which is pattern-drill). Good. Now, can we connect this to KNOWLEDGE?

	RR and RU	either NR or NU	total number of utterances
DATA 1	15 (39.5%)	23 (60.5%)	38 (100%)
DATA 2	3 (8.3%)	33 (91.7%)	36 (100%)

We can see that data 2 shows less reversibility. And because reversibility suggests external knowledge which is skills, we could misunderstand that there's more knowledge in data 2 than data 1. However, considering variety and degree of integration in different levels of knowledge, I believe data 1 is superior. In data 1, we could observe integration of an old knowledge and a new one, also integration of wording and meaning knowledge. Besides, there were different kinds of knowledge emerging though the exchanges. (sounding, wording, meaning) On the other hand, even if there shows externality than internalization in data 2, only wording level of knowledge were turned out to be emerging.

You can see that in the bold sentence above, Yura takes a bit of a logical LEAP. She begins to say things like "I believe" and "superior" without really explaining why she believes they are superior.

This is because she still have a gap to fill; she has shown us reversibility and non-reversibility, and she's also shown us various kinds of knowledge. But she hasn't really shown us how reversibility is related to knowledge.

Ⅳ. Discussion and Conclusion

From the comparison of two sets of data, we could see that there's more reversibility in a casual talk than in a grammar practice. I was starting with a hypothesis that if there's more reversibility the understanding of the teacher's utterances means that the children are getting more knowledge. But we have to consider 'internalization' and 'meta-talk' which are subordinate but significant factors. In data 1, we could find out that internalization doesn't always connect to the emergence of knowledge, rather, with intra-mental exchange(that is, externalization) produced more knowledge. Help! I'm lost here.

In data 2, we could see that when the teachers are using meta-language the reversibility declines. Good. But how does this answer our research question?

In a symmetric view of what?, I can make a conclusion that with casual talk, more various knowledge was shown and integration of different level of knowledge was possible compared to the data of a grammar practice. Yes, I think the qualitative analysis showed that. But how is the qualitative analysis linked to the quantitative one?

The internal validity of this conclusion is valid for the data under consideration, but I should say the external one is lower because of the different teaching-learning situations of all English teaching classrooms.

We saw earlier that Bibi took too many quotations out of the book, and she turned her back on the data.

Yura doesn't make that mistake. Her gaze is fixed firmly on her data, and she tells us a LOT about it, qualitatively and quantitatively.

But like Bibi, she's a little like a child in a candy shop. She's trying to pick EVERYTHING (reversiblity, three types, non-reversibility, three types, uptake, item-based combinations, three colors of that and five flavors of this).

You are going to get a stomach ache! Try looking at:

a) The INDEPENDENT variable: reversibility (functional, lexicogrammatical)

b) The DEPENDENT variable: predictabiilty (uptake, understanding)

I particularly like Yura's idea of predictability: it's a NEW idea, and I think it's a very good one. Let's develop that and forget about the old ideas in the book!

dk