APPENDIX B CHAT TRANSCRIPTION CODING SYSTEM

The conversations recorded for this study were transcribed and coded according to the conventions laid out in the Manual for Use of the CHAT Transcription Coding System of the Child Language Data Exchange System (MacWhinney, 1988). The edition used was that dated 24 March 1988. The reasons for the choice of this system are given in 3.4.5.a.

To facilitate the reading of the transcripts in Appendix B, a key to the main features of the CHAT coding system used in the transcripts is given below.

1. UTTERANCE DELIMITERS

The main line codes a single turn (3.4.6.b). The end of each utterance within a turn is marked by an utterance delimiter.

Period (Full Stop) .

A period marks the end of an unmarked (declarative) utterance. (Periods are eliminated from abbreviations.)

Question ?

The question mark indicates the end of a question.

Exclamation !

An exclamation point marks the end of an utterance which carries emphatic sentential stress.

Interruption [\]

This symbol is used for an utterance which is incomplete because another speaker begins talking or some event occurs causing the speaker to break off.

Self-interruption without retracing [///]

This symbol is used when a speaker starts to say something, stops, then continues on a completely different subject.

Trailing off [...]

The trailing off marker is the delimiter for an incomplete but not interrupted utterance.

2. PAUSES

[#] This symbol indicates a pause by the speaker. The length of the pause in seconds can be indicated as follows [# .5]

3. CODES FOR DIFFICULT MATERIAL

Unintelligible speech xxx

Only one xxx is used even if the transcriber can distinguish the number of unintelligible words or utterances.

Conversationally meaningful non-verbal turn O. (see 3.4.6.b)

Untranscribed material www

MacWhinney states that this symbol can be used when a speaker says something which has no relevance to the interactions taking place and which "the experimenter would rather ignore" (p. 14). In the current study it is used:

a) for a small number of instances where children play together for a short period with little or mainly unintelligible speech (eg between Turn 181 and Turn 1a2 in the transcripts of Child A).

b) for non-verbal turns which are not conversationally meaningful (3.4.6.b) but which are included because they provide information helpful in following the transcript (such as a change in location by one of the participants).

4. SCOPED SYMBOLS

These symbols are enclosed in square brackets and can be used to refer either to single words or to stretches of speech directly preceding the symbols. A stretch of speech referred to by a scoped symbol is enclosed in angle brackets. If no angle brackets are used, the scoped symbol refers only to the single word preceding it.

Quotation Mark ["]

This symbol marks the presence of a direct quotation by a speaker, or the performance of the role of a character. Its main role in the current study is to indicate that the speaker is talking for a puppet.

Example:

078*CO-: was that a good story? <yes. yes.>["] now you can go off to swing on the swing.

Guess in transcription [?]

Emphatic stress [!]

Overlap follows <text> [overlap>]

Overlap precedes <text> [<overlap]

Retracing without correction [/]

Retracing with correction [//]

Example:

037*COF: I was going to give him <a> [//] the hat. <I'll give him> [//]I'll show him the hat.

5. DEPENDENT TIERS

Dependent tiers are lines typed below the main line and contain additional information relevant to the main line. This information includes a number of codes relating to features such as intonation and gesture.

% par paralinguistic features.

This tier describes vocal qualities and vocalizations, such as laughing, crying and whispering. When a paralinguistic feature occurs at a specific point in the turn it can be placed in square brackets at that point.

Example:

033*BYF: a digger! [laughs]
034*BY-: [laughs] we had a digger [!].

If the paralinguistic feature occurs across the turn the relevant section should be enclosed in angle brackets and a separate tier should be used.

% int intonation and prosody

This tier describes qualities of pitch, amplitude and duration, rhythmicality and tempo.

%act actions

This tier describes people doing something.

%gpx gestures, proxemics

Gestures are actions (such as nodding, pointing or shrugging) which serve as language substitutes or accompaniments. Proxemics refers to the distance between speakers and also to the physical orientation of individuals with respect to other people or objects.

%add addressee

This tier indicates who is being spoken to.

% com comments by investigator/transcriber

In the current study this tier is used to record instances of unestablished referent.

6. SCOPING CONVENTIONS FOR DEPENDENT TIERS

Scoping symbols allow the dependent tier to refer to a specific utterance, phrase, word, segment, or any point on the main line.

<bef> refers to the point before the turn on the main line.

<aft> refers to the point after the turn on the main line.

<4> refers to the fourth word of the main line

<2-4> refers to the second to fourth word of the main line.

<1u> refers to the first utterance on the main line.

<2u4> refers to the fourth word of the second utterance of the main line.

<2u.> refers to the point after the second utterance on the main line.

<.2u> refers to the point before the second utterance on the main line.

<1u6-3u1> refers to the section from the sixth word of the first utterance to the first word of the third utterance.

7. INTERACTIONAL MARKERS

APPENDIX C TRANSCRIPTS

The transcripts will be provided as ASCII files in due course, linked from below.

APPENDIX D INTRODUCTORY QUESTIONS.

After watching a short video of the puppet, Louie, refusing to wear a hat and, on account of this, being refused permission to go on the swing, each subject was asked the following questions by the adult and, if necessary, provided with the answers.

QUESTIONS.

1. Does Louie know there's a swing outside?

2. Does he want to play on the swing?

3. Is he pleased with me?

4. Does he understand why he has to wear a hat?  

5. Can Louie see the hat?  

6. Would he be able to see it if he turned around?

7. Would it be a good idea for you to shout at Louie?

8. Why does Louie have to wear a hat?

APPENDIX E. PROTOCOL FOR THE ESTABLISHMENT OF REFERENT (P.E.R.)

SUGGESTED STRATEGY for identifying instances of failure to establish referent in a video-taped and transcribed conversation:

1. View the video to get an overall impression of situation, participants, interaction and any apparent breakdown or difficulty in communication.

2. Play through the video again, this time following the transcript and marking what appear to be any instances of failure to establish referent.

3. Go through the transcript and video in detail using the Substitution Test to check whether or not each referring expression has established referent for the addressee(s). (Substitution Test: Can you identify, and/or does the data give evidence that the addressee(s) could identify, the entity or set of entities indicated by each referring expression? Is the entity identified with appropriate specificity for the context?)

4. It is quite possible for all referents to be adequately established within an utterance which is inadequate in other respects (e.g. inaccurate, illogical, ungrammatical etc.). The sole criterion is whether or not the entity denoted by the referring expression is clearly established for the addressee(s).

5. Where ellipsis is involved the Substitution Test is applied to any referring expression(s) included in the omitted part of the sentence.

6. Requests for clarification can alert the coder to difficulties in the establishment of referent, but note that a contingent query is not necessarily a request for clarification. (For example the response 'You gave them what?' may convey amusement or disapproval rather than a failure to understand the referent indicated by 'what'.) The Substitution Test must always be applied.

7. In situations where there is more than one potential addresses it is necessary to determine who the addressee(s) is for every utterance. (There may be a shift of addressee within a turn.)

8. The analysis should cover all participants in the conversation.

CATEGORISATION OF REFERRING EXPRESSIONS:

The referring expressions identified by the above procedure are marked on the transcript and listed under turn number. Each expression is then categorised using the following headings.

A. Pronominal

B. Demonstrative

C. Definite Article

D. Comparison

Referring expressions in groups A-D (identity of reference) are further categorised as: Exophoric (reference to physical context) / General phoric (reference to social context) / Endophoric (reference to linguistic context) / or as examples of Discourse Deixis.

E. Substitution

F. Ellipsis

G. Lexical Item: a) meaning not understood b) inappropriate degree of specificity.

H. Identifying features: a) attributes b) locating devices c) literal comparisons d) other.

I. Adverbial of time.

J. Non-literal. (Includes metaphor, teasing, irony.)

K. Other.

In addition, for each referring expression the following questions are asked:

a. What is the entity referred to (state if guess only) and what type of entity is it? (person, object, location. time, action, idea, attribute etc.)

b. Clarification requested? Clarification attempted? Clarification successful?

Each instance of failure to establish referent is examined within the context of the conversation. The aim is to identify in what ways the processes necessary to the establishment of this referent have broken down. These processes include relevance to the conversational topic, an appropriate degree of specificity, and the taking into account of what is and what is not shared information with the conversational partner(s). It may be helpful to scan the transcript for any instances where there was potential for similar breakdown.