Language Models

Frameworks of some Models of Language.

What is typical of each form of language ?
What are the characteristics of each type ?
When you have identified the characteristics of these models, compare them with the features evident in your own chosen recording.
How close is your recording to the model ?
Now you should have a basis for comparison and contrast.

1. Unscripted conversation.

Non-fluency features are common and evident. There is repetition; there are pauses as hesitation, stuttering perhaps. This may confirm that the speech is spontaneous and not pre-planned. There may be evidence of self-monitoring , which may show itself as incomplete syntax for example. There may also be a change in the direction or topic of the conversation, a revision of something already said: "Well, I don't actually mean all teachers, but..." Turn taking procedures may reveal spontaneity - interruptions, overlaps, increases in speed and volume by the person as they attempt to continue "holding the floor". Equally the listener may attempt to interrupt by first grunting "mm" with an intonation which suggests doubt, disagreement or approval before entering the conversation at the first pause. "Yes but..." fulfils the same role - it opens by suggesting agreement and immediately moves on to contradiction. The opening word "yes" or "well" may be redundant in terms of meaning, though significant in terms of marking an entry into the conversation.

Supra segmental features can be indicators of spontaneity. They may be irregular - not following the pattern of meaning at all - or varying in speed, volume, pitch and stress in an apparently illogical fashion. Fillers are common. "You know" is a common device used to fill in gaps where another speaker might take over the conversation, and also to hold the floor while the speaker is searching for more to say.

Echoes are often used to show agreement and approval. To repeat the other speaker's words may be a sign that you agree with him. On the other hand echoes with rising intonation may suggest a doubt or question which hints at disagreement or misunderstanding. Either way it is characteristic of spontaneous conversation that there should be interaction between speakers. Syntax is likely to be complex by its irregularity. Utterances may become convoluted and lengthy, with sections incomplete. Pauses may be used as aural punctuation marks and may be linked with tone units but may also simply act as irregular breathing spaces between fast lengths of speech.
Clichés may be common in speech as a set phrase is more readily uttered than a totally new sequence. While clichés are rightly deplored as features of "sloppy" speech, the effort to speak in original phrases at all times would make speech arduous. Lexically, spontaneous speech may be impoverished compared with written speech - in which there has been time to consider a broader range of vocabulary and a more varied turn of phrase. There may, therefore, be an impression of informality in speech while written language with its more finely honed syntax and more complex vocabulary appears formal.

2. Scripted Interview

There are two aspects here which contrast with spontaneous speech.
The first is that, by virtue of being written initially there is time for revision and rewriting. The medium is different, the circumstances allow time for consideration, any errors can be reviewed and erased before speech takes place.
The second is that in an interview the speakers are not equal; one is in a position of dominance or control. That person also has the prepared notes while the interviewee is usually required to speak spontaneously. We would expect, therefore, to see a notable difference between the two speakers.
A scripted speech would be expected to be much more coherent than an unscripted speech. We would expect an absence of hesitations, repetitions and other normal non- fluency features.
Syntactically a scripted speech would be much more regular. Lexically it would be likely to be more varied. There would be no evidence of self monitoring, less evidence of dialect features (dialect is unusual in the written mode but common, even unconsciously, in speech).
Register is more formal. Supra segmental phonemes may be used in a regular and methodical way. It is not unusual for these to be written into speeches in order to bolster the effectiveness of the message.

The nature of an interview is that one person asks the questions while the other responds. This dictates the relationship of the speakers, the balance of speech and the character of the utterances. An interviewer may be aggressive (Robin Day to John Knott MP) or servile (Alastair Burnett to Prince Charles - parodied here in Spitting Image) but in both cases controls the progress of the conversation by asking questions and expecting answers.
A scripted question may be confident and coherent followed by an initial stumbling from the unprepared interviewee: "Ah, well, I suppose you could say that in a sense, however..." but a confident interviewee can confuse the issue by throwing a question back: "I believe that's an intrusive question. Let me ask you what you would do in that situation." This is, however, unusual and generally breaks the conventions of interviewing. Normally the interviewer asks briefly and the interviewee answers. It is noticeable however that experienced interviewees can dodge a direct question by answering obliquely. Faced with the direct question "We all know the government is aware that it is unpopular with farmers. What are you going to do about the price of pork?" An answer could be "No government can afford to ignore an important part of its electorate, but no government could contemplate unconditional subsidies." In other words, there is no direct answer to the question.

3. Prompted Speech

This is a way to describe that type of interview which appears to be informal and spontaneous but has as a framework, a pre-planned series of written questions or prompts. The interviewer may ask a scripted question, receive a spontaneous answer, add an unscripted comment then return to the script.

In this way the prompted interview exhibits some of the features of both spontaneous and scripted speech.