Discourse Analysis: The Sociolinguistic Analysis of Natural Language

by Michael Stubbs

Linguistics has traditionally concentrated on studying single sentences or isolated speech acts. In this book Michael Stubbs explores one of the most promising new directions in contemporary linguistics—the study of many sentences and how they fit together to form discourse. Using many examples drawn from recorded conversations, fieldwork observations, experimental data, and written texts, he discusses such questions as how far discourse structure is comparable to sentence structure; whether it is possible to talk of "well formed" discourse as one does of "grammatical" sentences; and whether the relation between question and answer in conversation is syntactic, semantic, or pragmatic.

Genres: Linguistics, Language, Nonfiction

279 pages, Paperback

First published January 1, 1983

Reviews

[Trevor · Rating 5 out of 5]

I wrote to a friend a few days ago that I was reading this book and could hardly put it down – she wrote back saying she thought it amusing that a book with such an abstract title would hold me enthralled – but look, it really has. I’ve been seeing much more of my intellectually disabled older sister lately and so have become aware of communication ‘mistakes’ she makes. Two of these relate to conversational turn taking. One is that she will jump in when she thinks of something she wants to say, no matter who else is talking or where they are up to in their part of the conversation. This is disconcerting, as what she has to say is rarely ‘on theme’. The other thing she does – and this shows my own bad nature – is that if you say something stupid to her (like asking her if there will be nude dancing at a party she is going to) she won’t answer the question, but she still takes her turn in the conversation – again with a completely random comment not related to the previous flow of the conversation.

The point is that most of linguistics assumes that there is very little that can be said about language over and above sentence level. Most of linguistics is concerned with whether or not a sentence is ‘well formed’. However, the ‘mistakes’ my sister makes have little or nothing to do with the internal grammar of her sentences. Sociolinguistics is very concerned about the rules that apply to language beyond the sentence level. For a sentence to be ‘well formed’ it really also needs to makes sense within the conversation and a sentence level analysis of the sentence is only part of the problem here.

A lot of this book discusses the methodological issues involved in studying language. One of the most interesting – that I hadn’t really thought about – was the problem of properly transcribing language so as to be able to analyse it. As he says, if you have a good quality recording, four minutes of conversation can take an hour to transcribe. This means you need to be selective about what you are going to analyse and being selective has its own issues. And then there is the observation paradox – where any observation involves someone being observed and being observed is pretty much the best way known to change the behaviour of the person being observed. And on we go down the rabbit-hole. There are a number of ways around these paradoxes – but the point repeatedly made here is that language use is so context dependent – that is, that we change the language we use to match the social requirements of context – that it is impossible to say what is us ‘being natural’. All communication involves, by definition, being observed – and so who is observing affects how we communicate. The point then is to see how particular social situations impact on language use.

One of the things I found particularly interesting in this was how much of our spoken language in particular is spent ensuring that we are on the same wavelength as those we are speaking with. He even suggests that spoken language is heavily dependent on adverbs and adverbial phrases mostly because these help keep people on track.

But the other interesting thing is the exact opposite tendency in communication – and that is that often the literal connection between statements is very hard to grasp. I’m not just talking about idiomatic expressions (although, obviously, these too are socially bound) but sometimes even our simplest statements can seem, on a close and logical analysis, to be nearly random (even when they are anything but).

One of my favourite examples is of a teacher saying to a group of students – ‘right, fags out’ – I’m not sure how that translates into American English, but essentially, ‘come on, students, time to put out your cigarettes’. Now, obviously, none of the students are smoking. The point is that to understand that this statement actually means, ‘we are coming back together as a group now – so, you all need to stop what you are doing and pay attention to me’ is anything but clear from the literal meaning of the words. And it only makes sense in a culture in which socially, even amongst those supposedly too young to smoke, a break is often referred to as a ‘smoko’.

There is also a lovely bit of this where he is discussing the differences between logic and language. He does this with an extended discussion on ‘and’. In many ways we like to think of ‘and’ as being like + in maths. You know, ‘x+y=y+x’. Except that ‘and’ really doesn’t work like that. As he explains, ‘I crashed the car and got drunk’ simply doesn’t mean the same thing as ‘I got drunk and crashed the car’.

The point of this book is to show that there are linguistic structures that we can meaningfully talk about that occur above the sentence level and that there are rules we can follow (and do follow) in constructing conversations which are more important than many of the supposed rules of sentence construction. We can say more about narratives than merely what genre they belong to – but, even here, being able to talk about genres at all probably makes his point. There is a wonderful analysis of a short conversation between two Scottish school children and the author about a haunted house and a fascinating analysis of Hemmingway’s Cat in the Rain. In fact, his analysis of Hemmingway’s short story was really interesting. He got students to make two summaries of the story – one under 60 words and the other under 25 words – and then he analysed the two versions they produced. Interestingly, the story is about problems within a marriage and possibly also an Italian hotel owner who fancies an American woman (none of which is literally said in the story) – but the shorter the summaries, the more likely people become literal. That is, the more likely people talk about cats and rain. I thought this was very interesting. If I end up teaching next year I can see me doing this exercise quite a few times.

I don’t know if this review has convinced you of why I found this book quite so hard to put down – but I did enjoy the book, I enjoyed it very much.

[Atif]

how i can read this book?