THE METHOD OF SERIAL REPRODUCTION

The sign language transcriptions were made by the author's informant, in conjunction with the author. Capital letters are employed in the conventional way for T and sentence initially. Elsewhere they indicate fingerspelling (and this occurred quite frequently, sentence initially – presumably reflecting the influence of orthography on sign language). The main feature of the transcripts is the use of ( ) brackets. Any word, words or phrase enclosed in ( ) is either a comment or description, or it is a sign which is not accompanied by a lipreadable mouth movement. This is apparently a characteristic of Southern English Sign Language.

Words in italics indicate a lipreadable mouth movement which is not accompanied by a sign. Superscripts indicate signs with a different form, and other devices used are [idiom] and |pause|.

Punctuation is inserted where it seems clear from the tapes, not where it would seem appropriate from the translation. The purpose of the “direct transcripts” is to record, as closely as possible, what is found in the sign language on tape.

The translations are to be considered a secondary, or even tertiary, form of the data. They are included in response to comments and criticism received at the Symposium. They make clear the manner in which the Story Chains develop. They were done by the author's informant (in conjunction with the author) direct from the tapes and with only occasional reference to the transcripts.

These experiments are based on a very similar technique which exists probably in many cultures but certainly in the English speaking world as a children's game. For this the children arrange themselves so that in turn they can have a message whispered into one ear, which they then repeat into the ear of the next child, and so on.

At the end the last child tells his version aloud and all is merriment as the various versions are compared with the original. This procedure – the imperfect passing of an unwritten message from one person to the пехt – is called the Story Chain, and it is a very rich experimental paradigm. It offers a very natural way of studying several aspects of the process of communication – memory, perception, variation and some of the differences between the use of language for dialogue and monologue. The game as described does, however, need some modification if the study of memory and perception are not of prime concern.

The Story Chain procedure which was actually employed did not require the participants to misperceive or misunderstand, they were merely required to pass on the message as faithfully as possible. With hearing children the success of the game depends upon misperception induced by the whispering. However, with deaf subjects the presumed idiosyncracy and interpretative-ness in communication behaviour is considered likely to induce sufficient error to ensure that procedure resembles the children's game. It was assumed that this prodedure would be sufficiently motivating to ensure a good sense of participation in something “game-like” - that, in fact, the subjects would find the task interesting and good fun. The procedure does have its drawbacks however, and it is appropriate at this juncture to consider the techniques in a little more detail.

In the communication environment under discussion we have to consider that variation limiting is taking place, that is, that the language models of deaf children are sufficiently different one from another for communication I to be just tolerably unsuccessful and critically dependent on interpretative I skills. Failure must be expected and unsurprising (that is, in normal communication experience) and it seems likely that this is due in part to grammatical differences as well as semantic and lexical difficulties.

One question, therefore, which we should address is how can the Story Chain be used to detect idiosyncratic grammar? We could proceed to answer that by assuming that the Story Chain data will contain sufficiently complex I misunderstandings which leave one with no choice but to assume the idiosyncracy is there. Another possibility which seems to be worth exploring arises from consideration of how a dialogue would proceed if language were indeed such an individual thing. The conjecture is that dialogue in sign I language is likely to involve negotiation of grammar as well as lexicon and I therefore that communicative success will have to depend on frequent monitoring of the addressee's understanding.

The experimental implication of this is that the ability of a storyteller to monitor his addressee's understanding might be open to experimental influence through alteration of the Story Chain procedure. Specifically, if some sort of monologue condition could be contrived (and there are two possibilities for this) then some of the processes of negotiating understanding could be examined. In fact, this possibility was explored experimentally, although space limitations prevent more than a passing mention.

(W. Edmondson . The Story Chain. – London, Language in Sign an International Perspective on Sign Langual, 1983.)