Re:data analysis based on Tomasello's three categories

[kellogg]

I'm going to code each utterance like below (but I had to struggle to define which turn should be included in which category) :

Good--can you give me an example? I might be able to help!

① The fixed EXPRESSION‎!S → EX

② The item-based COMBINATIONS → CM

③ The abstract CONSTRUCTIONS → CO

④ Non-expression‎!s, Non-combinations, and Non constructions → NON

▶Data: Jisu's 6th grade data of the first class of lesson 2, "Is This York Street?"

T: OK. Open your books to page 16.

Look at the picture on page 16 and I have a question.

Please raise your hand if you can answer.

Who do you see in the picture?

Who do you see in the picture?

(Ss raise hands) 어...민지?

민지: I can see grandma? CO

T: Uh huh~

민지: and some of family? CO

T: Mmhmm~

민지: and building. CO

T: And buildings?

OK, be seated.

OK. when I ask, "who do you see in the picture?", 민지 says "I can see~"

That's really good answer.(sic) "I see," or "I can see..." (pointing to Kevin in the picture)

Who is he?

Ss: !@#%$! (inaudible) NON

T: Who is he?

You know who he is.

Who is he?

Ss: Kevin. CM

But this appears, word for word, in the lesson. Doesn't it?

T: Yes, he is Kevin.

And... who are these?

S1:...... Families? CM

This is an error. We don't use "families' to mean "family members". It's a mass noun. Since you coded "and building' as a construction on the basis of the error, shouldn't you do the same here?

T: We don't know. (at the same time)

T: 몰라요.

These foreign people, foreigners, visitors.

And she(pointing to 민지) says she is grandma.

But is she,, 민지‘s grandma?

Ss: No. CM

"No" appears many times in the text. Besides, it has no internal structure. How can it be a combination? I don't understand how you are coding this.

Remember that the way you code data depends on the research question you are asking, Seong-eun. Let's have a look at your research question and maybe we can straigten this out.

. 

1) A noun:

  LEARNER TALK - abstract creative constructions

2) A wh-question:

  How does the proportion of fixed expression‎!s, item-based combinations, and  abstract constructions CHANGE as the term from September 2009 to December 2009, from Lesson 9 to Lesson 16 ?

This suggests that you ARE interested in how the fixed expressions ACCUMULATE. So you SHOULD include ALL the previous material the children have studied as "fixed expressions". Right?

  How does the teacher get the learners to create longer and more complex constructions? What kinds of teacher production contribute to learners' long and complex language constructions?

This suggests a MICRO-genetic focus. Good!

Try looking at the teacher talk that comes BEFORE a "long complex" learner turn.   

  Does the proportion of fixed expression‎!s, item-based combinations, and abstract constructions to total utterances go up or down?

I'm not clear--are you taking about September to December? Or are you talking about in a lesson?

 Does the teacher production (more open questions rather than closed questions, talk about more INFORMATION rather than THINGS, more reference to GOODS and SERVICES rather than LANGUAGE, more symbolic meaning rather than iconic meaning, more talk about words and rules rather than indication of pictures and objects, etc.) significantly affect the learner production of abstract constructions?

OK--now you are talking about the single lesson again. Try putting the MACROGENETIC questions (September to DEcember) together and the MICROGENETIC questions (teacher to learner) together. It's very confusing to be going back and forth like this.

Now, for the MACROGENETIC questions you need proportions. And they have to be pretty big.

For the MICROGENETIC questions you need to look at UPTAKE--that is, how many of the TEACHER'S words to the kids use in a construction. So we might want to have a DIFFERENT definition for the two timescales:

MACROGENETIC: uptake from the book (fixed expressions from the book, abstract constructions on the basis of fixed expressions found in the book, and abstract constructions based on the structures found in the book)

MICROGENETIC: uptake from the teacher talk (repetition of what the teacher says, variationi of something the teacher says, completely original answer.

T: No.

So that's the difference between Korean and English.

우리가 할머니라고 얘기했을 때 보통 할머니 그러면 우리가 영어로 말하면 grandmother. 근데 우리가 말하는 할머니는 우리 할머니도 되고 그냥 지나가는 할머니도 되잖아. 근데 영어에서 grandma, grandmother 그러면 나의 할머니가 되는 거야.

Then, how would you explain this woman in English?

Can you think of a good way?

Not grandma...

What do you think ? (시형 raises his hand) 시형?

시형: !@#%!%

T: (raising a hand towards her ear)

시형: Old woman. CO

Good.

T: Old woman?

(덕환 raises his hand) What do you think, 덕환?

덕환: I think !@#$!% (inaudible, I don't remember) CO

OK.

T: OK.

That's one way to say that, but old woman is not a nice way to say that.

S1: Madam. CO

Actually, a form of this does occur in the song of the lesson  "Oh, well excuse me ma'am..."

It's interesting that the teacher doesn't seem to know this.

T: Madam?

Could be?

자, 여기서 old woman 이라고 하면

Ss: 늙은이..(laughing) NON (because it's a Korean word)

T: 늙은이가 되잖아. 그치? 누가 자기보고 old 라고 했을 때 기분이 별로 안 좋겠죠? So 그래서 old lady 라고 하면 알아는 듣겠지만 좋은 말은 아니야. 이럴 때 우리가 영어표현으로 쓸 수 있는 게 이런 중년의 여성이라는 뜻으로 어떤 표현을 쓸 수 있냐면 ..(writing " a midd....)

S1: a middle-aged woman. NON (because S1 just read it on the board)

But then ALL uptake has to be considered NON? That doesn't seem right. 

T: Yes, a middle-aged woman이라는 표현이 좀 더 나아요.

So we have Kevin, two foreign people and one middle-aged woman, what else do you see in the picture?

What else do you see?

성해, what else do you see?

I see...?

성해: I see...NON (because 성해 just repeated T's initiate)

But then ALL uptake has to be considered NON? That doesn't seem right.

Remember the topic, Seong-eun:

LEARNER TALK - abstract creative constructions

Isn't it possible that SOME abstract creative constructions (maybe even MOST of them) will be created from teacher utterances?

T: Uh-huh~

성해: Apple street? CM

This appears word for word in the dialogue, of course. Isn't it a fixed expression‎? 

T: Uh-huh~

You can see Apple street.

OK, what's in the balloon?

성해: 어,,,Apple street,, (someone seems to have said something to her) CM

See above.

I can see a map. CO

There are two possibilities:

a) This is created from "I can see grandma". It's a microgenetic item-based combination, because it involves uptaking an utterance by another learner and chaning it in exactly one place.

b) This is created from the teacher. It's a microgenetic abstract construction because it involves manipulating abstract rules of grammar to create an indicative declarative form from an interrogative (OR: it is a microgenetic abstract construction because it involves changing the teacher's utterance "You (-->I) can see Apple Street (-->a map)."

T: Yes, you can see a map.

So, what's on the map?

What's on the map?

(Ss raise hands) 음~ 신미?

신미: Apple street, York street, bus stop, post office, Korean restaurant. CO

Isn't this microgenetic uptake from the written material?

T: Yes, Apple street, York street, post office, Korean restaurant and bus stop.

So, what do you think they are doing?

What do you think they are doing?

(Ss raise hands)하영?

하영: They're asking, what do we %$#@#% (inaudible), uh, they're asking,..........CO

T: They are asking,,,,maybe.. they are asking and answering about...

하영: About the,,,,where do they want to go...? CO

Wow! reported speech. Now, it's clearly 'scaffolded" by the teacher. But this is an example of exactly what you are looking for in your microgenetic research question. So you need to look VERY carefully at the teacher talk. What is it about the teacher talk that makes this precocious use of reported speech possible?

T: Uh-huh, good job.

(시형 raises his hand) 시형, What do you think they're doing?

시형: They are asking about directions. CO    

And this is really the very FIRST CO which has no errors.

T: Directions?

Maybe. Right.

So, you are going to watch a video clip.

Eyes on me.

Pencils down. Pencils down.

Eyes on me. 태영? (he was being distracted)

Thank you.

So you are going to watch a video clip.

Before you watch it, I have two questions for you.

So be prepared for the questions.

My first question is, "how do you get to York Street?"

입균 (he was being distracted too), What is my first question?

입균: !@#?NON

a) What is the QUANTITATIVE relationship between expression‎!s, combinations, and constructions? Which is proportionally greatest? Compare it with our 3,4, and 5th grade data.

Form	Turns	Percentage
Expression‎!	0 I found one.	0.0
Combination	4	20.0
Construction	11	55.0
Non-ECC	5	25.0
Total	20	100.0

As can be seen, I didn't code the whole data though, any fixed expression‎!s(0%) from learner talk is not shown from the data, plus, only 20.0% of item-based combinations. Instead, we can see the greatest proportion of abstract constructions (55.0%) as contrasted with that of 3 (8.7%), 4 (14.7%), and 5th (13.1%) grade. To compare the result with our 3, 4, and 5th grade data, while the fixed expression‎!s proportionally increase from third (47.1%) to fourth (54.6%) to fifth (68.1%), they unbelievably disappeared in the 6th grade data above. That must be because the student talk is produced from "Look and listen" talking about the picture to guess what will happen in the story and what the students are supposed to listen carefully and learn, that is, teaching point, before Ss really get down to the process of learning the target language, or before they do not learn the fixed expression‎!s by rote yet. (Besides the children in the data seem to be very advanced in language development. They must be given much freedom to speak out what they have learned in the private institutes or overseas.) 

Take a look at Hayeong's data, Seong-eun.

 Now, we will present the result for constructivity of students' production through different activities, according to Tomasello's Categories:

Constructivity   

Activity	Mean	N	Standard Deviation
Greetings	1.00	41	.63
Chat	1.02	50	.38
Listen and Do	.65	43	.69
Look and Listen	1.06	162	.81
Listen and Repeat	1.08	111	.56
Listen and Answer	1.09	126	.84
Total	1.03	533	.72

We will explain this result with extended discussion in the next section.

As you can see she found that there quite a bit of construcion in Look and Listen. It's the third most constructive activity, well above the average (.103).

But it's actually LESS constructive than Listen and Repeat. This is mysterious!

I'm not saying it's right. I'm not saying it's wrong. I'm saying it's a mystery. Hint: as you know, there were some pretty serious problems with Hayeong's data!

b) Compare teacher talk with student talk. Which has the highest proportion of constructions?

Teacher talk in the data is mostly (more than 70%) consist of metatalk, that is, the teacher's utterances/turns are seen very long and complex, Careful! That's NOT what meta-talk means. as professor Kellogg mentioned, metatalk is almost always abstract constructions.

Meta-talk is talk about talk. For example, when you say "Listen!" or "Repeat!" this is metatalk. It doesn't appear in the dialogue usually, so it's quite likely to be an abstract construction, though of course many teachers use them so often that they become fixed expressions.

 So the answer of this question could be that teacher talk has the highest proportion of constructions. I can notice that the higher the proportion of constructions of teacher's metatalk appears, the less proportion of creative student's constructions we get.

Really? Where?

It's a good hypothesis, Seong-eun. But how can we prove it?

c) Find the longest and most complex turns. Find the longest and most complex exchange. Describe them qualitatively.

  The longest and most complex turn maybe 하영‘s turn, " About the,,,,where do they want to go...?" and 시형’s turn, "They are asking about directions." 하영 seems to be trying to construct a compound complex sentence like, "They are asking and answering about where they want to go/get", which contain more than two verbs and more than one subordinations. Although 하영 failed to construct a complex and grammatically correct sentence, she is predicted to be in the process of language development. But 시형‘s long, complex, and grammatically correct sentence, "They are asking about directions" shows us that he is combining all the units (the present progressive form, the preposition, and the word knowledge of 'directions', etc) to construct a perfect language with his previous knowledge. He can be said to proceed toward the next stage of development.

Bravo, Seong-eun. But mind the gap! First of all,  before we talk about the child's development, let's talk about the LEARNING that is leading it and let's not assume that ALL learning leads to development.

Secondly, let's remember the Assistance and Potential assumptions.  Look at what the TEACHER says. Remember that Jisu is an EXPERT. She does something quite artful, quite expert. The beauty of this particular construction is not reducible to either Jisu's assistance or to  시행's immanent potential.

  The longest and most complex exchange could be the part about 'a middle-aged woman', which still doesn't seem to be socioculturally correct though. Students and the teacher show that they are struggling to construct a new abstract concept for 할머니 which means grandma as well as old ladies in Korea, but it has different concept in English. To build a new sociocultural knowledge or lexical concept, students try to use all of their previous knowledge with the teacher's assistance progressing toward development.

Good! I'm not quite sure what to make of this. Jisu says that she wants to avoid scientific concepts. But of course the concept she is trying to express here, which I would certainly qualify as ethnographic, is a scientific one, in the sense that it is NOT accessible through everyday life.