Send the link below via email or IMCopy
Present to your audienceStart remote presentation
- Invited audience members will follow you as you navigate and present
- People invited to a presentation do not need a Prezi account
- This link expires 10 minutes after you close the presentation
- A maximum of 30 users can follow your presentation
- Learn more about this feature in our knowledge base article
Transcript of 123
The best way to understand the meaning of discourse is through comparing it with the notion of text
There are two different types of
(the language that surrounds or accompanies the piece of discourse under analysis)
(experiential context within which the discourse takes place)
by David Nunan
What is discourse?
1. "discourse'': A continuous stretch of (especially spoken) language larger than a sentence, often constituting a coherent unit, such as a sermon, argument, joke or narrative.’ (Crystal 1992: 25)
2. "text": A piece of naturally occurring spoken, written, or signed discourse identified for purposes of analysis. It is often a language unit with a definable communicative function, such as a conversa¬tion, a poster,’ (Crystal 1992: 72)
4. "discourse": stretches of language perceived to be meaningful, unified, and purposive.’ (Cook 1989: 156)
5. "text": a stretch of language interpreted formally, without context. (Cook 1989: 158)
the terms ‘text’ and ‘discourse’ are interchangeable
discourse analysis involves the study of language in use
a text or piece of discourse consists of more than one sentence and the sentences combine to form a meaningful whole
Non-linguistic contexts include:
the type of communicative event (for example, joke, story, lecture, greeting, conversation);
the purpose of the event;
the setting, including location, time of day, season of year and physical aspects of the situation (for example, size of room, arrangement of furniture);
the participants and the relationships between them;
the background knowledge and assumptions underlying the com¬municative event.
Spoken versus written language
Written language does, in fact, perform a similar range of broad functions to those performed by spoken language - that is, it is used to get things done, to provide information and to entertain. However, the contexts for using written language are very different from those in which spoken language is used.
suggests that written language is used
Linguistically, written language tends to consist of clauses that are internally complex, whereas with spoken language the complexity exists in the ways in which clauses are joined together.
The written text usually seems to have more information packed into it.
Spoken and written language also differ in the ratio of content words to grammatical or function words. The number of lexical or content words per clause is referred to as lexical density.
The density of written language is also reinforced by the tendency to create nouns from verbs. Examples of this process arc as follows:
Good writers reflect on what they write.
Reflection is a characteristic of good writers.
calls this process of turning verbs into nouns grammatical metaphor.
The words themselves must carry all of the shades of meaning which in face-to-face interaction can be conveyed by non-verbal behaviour.
Written language is more densely packed with information than spoken language.
There is no opportunity for the readers to signal that they do not understand.
• The spoken text accompanies a sequence of actions in context, whereas the written text is written for an unknown audience which is distant in time and place.
•In written text, there are numerous references to things outside of the text itself. These are not explicitly identified but are referred to by words and phrases such as this, it, that, this one, that one, it, etc. In written texts, because the writer must construct a context for the interaction, objects, people, actions etc. are named.
With written language, there is no common situation, as there is in face-to-face interaction. The situation therefore has to be inferred from the text.
The texts could be divided up is into dialogue and monologue.
Another division would be into those that are basically transactional in nature, and those that are basically interpersonal.
Transactional language is that which occurs when the participants are concerned with the exchange of goods and services.
Interpersonal language, on the other hand, occurs when the speakers are less concerned with the exchange of goods and services, than with socializing.
Types of discourse
This distinction — between language which is used to get goods and services, and language which is used to fulfill a social purpose - is a common one in the literature. Many interactions that are essentially transactional in nature will also exhibit social functions, while essentially social interactions can contain transactional elements. Some conversations have both transactional and interpersonal functions.
In some texts, the primary purpose is an expressive or aesthetic function. This aesthetic function is a third major purpose for which people use language
- that is, sequences of sentences or utterances which seem to ‘hang together’ - contain what are called text-forming devices. These are words and phrases which enable the writer or speaker to establish relationships across sentence or utterance boundaries, and which help to tie the sen¬tences in a text together.
Halliday and Hasan identified five different types of cohesion:
reference, substitution, ellipsis, conjunction and lexical cohesion
. In Halliday (1985a) these have been further refined and the five categories have been reduced to four, with substitution being seen as a sub-category of ellipsis.
If a single sentence is taken out of context and presented in isolation, it is likely to contain elements that are difficult, if not impossible, to interpret.
Comparative reference is expressed through adjectives and adverbs and serves to compare items within a text in terms of identity or similarity.
A: Would you like these seats?
B: No. as a matter of fact, I'd like the other seats.
Demonstrative reference is expressed through determiners and adverbs. These items can represent a single word or phrase, or much longer chunks of text - ranging across several paragraphs or even several pages.
"Recognizing that his country had to change, Gorbachev could have become a cautious modernizer in the Chinese fashion, promoting economic reform and sponsoring new technology while holding firm against political change. This did not happen."
Personal reference items, are expressed through pronouns and determiners. They serve to identify individuals and objects that are named at some other point in the text.
"Mikhail Gorbachev didn't have to change the world. He could have chosen to rule much as his predecessors did."
Substitution and ellipsis
There are some new tennis balls in the bag. These ones've lost their bounce.
A: Annie says you drink too much. B: So do you!
A: Is it going to rain? B: I think so.
In each of these examples, part of the preceding text has been replaced by ones, do, and so respectively.
Ellipsis occurs when some essential structural element is omitted from a sentence or clause and can only be recovered by referring to an element in the preceding text.
My kids play an awful lot of sport. Both (0) are incredibly energetic.
A: Have you been, working? B: Yes. I have (0).
A: Why'd you only set three places? Paul's staying for dinner, isn’t he? 3: Is he? He didn tell me (0).
Halliday and Hasan point out that these two types of cohesion are essentially the same.
It is a cohesive device because it signals relationships that can only be fully understood through reference to other parts of the text.
I afraid I'll be home late tonight.
, I won’t have to go in until late tomorrow.
From a marketing viewpoint, the popular tabloid encourages the reader to read the whole page instead of choosing stories.
isn't that what any publisher wants?
: Brick tea is a blend that has been compressed into a cake. It is taken mainly by the minority groups in China.
, it is ground to a dust.
it is usually cooked in milk.
:Chinese tea is becoming increasingly popular in coffee shops. This is
of the several health-giving properties.
'Given' and 'new' information
In English, there is a ‘
’ word order of
Subject + Verb + Object
. The notion of a ‘standard’ word order has some sort of psycholinguistic as well as grammatical plausibility.
The cat ate the rat
' - is a sentence that exhibits the standard Subject + Verb + Object structure. However, there are numerous other ways in which the semantic content of the sentence could be expressed. One important consideration is whether the information has already been introduced into the discourse, or is assumed to be known to the reader or listener.
Such information is referred to as
. It contrasts with information which is introduced for the first time and which is known as
It is the speaker/writer who decides what information should be considered given or new.
We can see the close relationship between discourse considerations and grammatical structuring in relation to given and new information.
is a formal grammatical category which refers to the initial element in a clause. It is the element around which the sentence is organized, and the one to which the writer wishes to give prominence.
Theme and rheme
Everything that follows the theme is known as the
Within the school of linguistics known as
functional linguistics, three types of
are identified -
themes have to do with the information conveyed in the discourse.
themes, on the other hand, reveal something of the attitude of
the speaker or reader.
themes link a clause to the rest of the discourse.
When moving beyond the sentence to discourse, the issue of
becomes particularly important as the writer has to arrange information in terms of given/new and also in terms of desired thematic prominence.
In recent times, the term ‘
’ has been adapted by functional linguists to refer to different types of communicative events (Martin 1984; Swales 1990). They argue that language exists to fulfill certain functions and that these functions will determine the overall shape or
of the discourse.
Different types of communicative events result in different types of discourse, and each of these will have its own distinctive characteristics. Some events result in sermons, others in political speeches, and yet others in casual conversations. While each sermon, political speech and casual conversation will be different, each discourse type will share certain characteristics which will set it apart from other discourse types.
A: What did you do last night ?
B; Well, Mum and Dad went out so we went to Marg ’s to sleep, and Sarah wouldn't go to sleep, and she wanted to ring Mum, and Marg said she couldn't, and so she cried, and so Marg combed her hair, and then she went to sleep. She was really naughty.. .
A: What time did she go to sleep?
B: Mmm - 'bout one o 'clock.
Each of these texts is very different in terms of its structure, grammar and physical appearance. The first text is an extract from a leading article in a newspaper. Its generic structure is as follows: title, author(s), location, argument, supporting details. In terms of its layout and physical appearance, the text contains a large, eye-catching headline. The columns and assignment of each sentence to a separate paragraph are designed to make the piece easy to read.
The second extract, is taken from a conversation between a girl and her grandmother, and contains a
. According to functional linguists, a
consists of a sequence of events which are initiated by an introduction and orientation, and which end with a comment and conclusion. Grammatically, recounts are characterized by the simple past tense, and the use of specific reference to people and places.
The systematic relationship between language structure and function is described by
in the following way:
Every text - that is, everything that is said or written - unfolds in some context of use; furthermore, it is the uses of language that, over tens of thousands of generations, have shaped the system. Language has evolved to satisfy human needs; and the way it is organised is functional with respect to those needs - it is not arbitrary. A functional grammar is essentially a ‘natural' grammar, in the sense that everything in it can be explained, ultimately, by reference to how language is used.
(1983) argues that the ordering of information in discourse can reflect certain rhetorical relationships such as cause-consequence, problem-solution. He uses the following four sentences to illustrate the ways in which these relationships function in discourse.
I opened fire.
I was on sentry duty.
I beat off the attack.
I saw the enemy approaching.
These four sentences can be sequenced in twenty-four different ways. However, not all of these sequences will be acceptable as coherent discourse. According to Hoey, only one sequence is completely acceptable:
I was on sentry duty. I saw the enemy approaching. I opened fire. I beat off the attack.
There are in fact grammatical devices that can be employed to change the sequencing of the information in the text in acceptable ways. These include:
(While I was on sentry duty, I opened fire, because I saw the enemy approaching. I (thereby) beat off the attack.)
(I opened fire because I saw the enemy approaching when I was on sentry duty. By this means I beat off the attack.)
is a single statement about some entity or event. A sentence may contain a single proposition or several propositions. Propositional analysis enables the researcher to compare texts that would not otherwise be comparable.
Not only are language factors involved in the processing of discourse, but the background knowledge and interests of the reader or listener have an important bearing on what is recalled. Psychologists have investigated the hypothesis that, all other things being equal, the difficulty of a text will be determined by the number of propositions it contains. In an experiment designed to test this hypothesis,
Kintsch and Keenan
(1973) predicted that sentence 1 would be more quickly read and understood than sentence 2.
1. Romulus, the legendary founder of Rome, took the women of the Sabine by force.
2. Cleopatra’s downfall lay in her foolish trust in the fickle political figures of the Roman world.
In addition to the discovery that the number of propositions in a text will have an important bearing on how easily a reader or listener will understand it, researchers have also found that there is a hierarchical relationship among the propositions in a text. Subsequent research has shown that the more central a proposition is to the argument of the writer, the more likely it is to be recalled by a reader.
Lexical cohesion occurs when two words in a text are semantically related in some way - in other words, they are related in terms of their meaning. In
, the two major categories of lexical cohesion are:
The logical connections that readers or listeners perceive in a written or oral text.
Cohesion is neither necessary nor sufficient for the creation of coherent discourse.
He goes on to suggest that we are able to recognize this text as coherent by creating a context and then identifying the functions that each utterance fulfills within that context. Most native speakers would create a domestic situational context in which the following functions are assigned to each utterance:
A: That's the telephone. REQUEST
B: I'm in the bath. EXCUSE
A: OK. ACCEPTANCE OF EXCUSES
In creating a meaningful context and identifying the functions of each utterance, coherence is established. As a result, the missing bits of conversation, which would make it cohesive as well as coherent, could be restored.
also explores the issue of what distinguishes text from non-text (that is, coherent from non-coherent texts). He claims that it is difficult to create non-texts from random sentences, because some sort of context can generally be created which will give coherence to any set of sentences.
claims to be in basic agreement with the notion that cohesion does not ‘create’ coherence. However, he believes that Widdowson and Edmonson overstate their case. Their argument assumes that each utterance has a clearly identifiable function, the perception of which is somehow independent of the ideas or propositions expressed.
The issue of which comes first - perception of the full propositional meaning, or the function performed by each utterance - is a ‘chicken and egg’ argument. In this particular instance,
believes that perception of the propositions must either precede or occur simultaneously with the recognition of functions.
Cohesive links between utterances were insufficient to account for the coherence of discourse, and that such coherence depends on the ability of the language users to recognize the functional role being played by different utterances within the discourse.
includes repetition, synonym or near synonym, superordinate, and general word.Reiteration thus fulfils a similar semantic function to cohesive reference.
can cause major problems for discourse analysis because it includes all those items in a text that are semantically related. In some cases this makes it difficult to decide for certain whether a cohesive relationship exists or not.
Many lexical relationships are text- as well as context-bound. This means that words and phrases that are related in one text may not be related in another. The background knowledge of the reader or listener plays a more obvious role in the perception of lexical relationships than in the perception of other types of cohesion. Collocational patterns, for example, will only be perceived by someone who knows something about the subject at hand. Text- or context-free word classes can only ever be partial, and the test of whether many items are cohesive or not will be determined by the particular text in which they occur. In addition, our ability to identify collocational relation¬ships in a text will depend on our background knowledge - that is, on our familiarity with the content of a text.
Hoey (1991) argues that lexical cohesion is the single most important form of cohesion, accounting for something like forty per cent of cohesive ties in texts.
Many lexical relationships are
text- as well as context-bound
. This means that words and phrases that are related in one text may not be related in another. The
of the reader or listener plays a more obvious role in the perception of lexical relationships than in the perception of other types of cohesion. Collocational patterns, for example, will only be perceived by someone who knows something about the subject at hand. Text- or context-free word classes can only ever be partial, and the test of whether many items are cohesive or not will be determined by the particular text in which they occur. In addition, our ability to identify collocational relationships in a text will depend on our background knowledge - that is, on our familiarity with the content of a text.
One well-known theory is frame theory. This suggests that human memory consists of sets of stereotypical situations, or ‘frames’, which are constructed out of our past experiences. These provide a framework which we use to make sense of new experiences. For example, our former experiences of ‘going to the doctor’ provide us with a frame that enables us to predict what is likely to occur when we next visit the doctor. One major problem is that it provides no explanation of why one frame might be selected rather than another.
Schema theory suggests that the knowledge we carry around in our heads is organized into interrelated patterns. These are constructed from all our previous experiences and they enable us to make predictions about future experience. Given the fact that making sense of discourse is a process of using both our linguistic knowledge and also our content knowledge, these schemata or ‘mental film scripts’ are extremely important.
The central insight provided by researchers using mental models such as frame and schema theory is that meaning does not come neatly prepackaged in aural and written texts. Widdowson (1978) has suggested that texts are little more than elaborate ‘signposts’ to the speaker or writer’s original meanings, and that the reader or listener must use his or her linguistic and content knowledge to reconstruct the original meanings of the creator of the discourse.
In a later work, Widdowson (1983) provides a novel reinterpretation of schema theory from the perspective of discourse comprehension. He argues that there are two dimensions of levels to any given discourse - a systemic level and a schematic level. The systemic level includes the reader or listener’s linguistic knowledge, while the schematic level relates to background content knowledge.
The things we know about the world assist us in the interpretation of discourse.
If the balloons popped, the sound wouldn't be able to carry since everything would be too far away from the correct floor. A closed window would prevent the sound from carrying, since most buildings tend to be well insulated. Since the whole operation depends on a steady flow of electricity, a break in the middle of the wire would also cause problems. Of course, the fellow could shout, but the human voice is not loud enough to carry that far. An additional problem is that a wire could break on the instrument. Then there could be no accompaniment to the message. It is clear that the best situation would involve less distance. Then there would be fewer potential problems. With face-to-face contact, the least number of things could go wrong.
Most native speakers have no trouble comprehending the grammatical structures and vocabulary items in this story. Despite this, they have a great deal of trouble understanding what the text is all about, and even greater difficulty in providing an oral or written summary.
The passage is from a well-known study by Bransford and Johnson (1972) which demonstrated the importance of context and background information for the interpretation of discourse. They found that subjects who were asked to listen to the text and recall it had a great deal of difficulty. However, another group of subjects who were provided with a picture were able to recall virtually all of the text.
This interaction between the world of the text and the world outside the text is exploited by writers in many different ways. For example, some writers often create humorous or satirical effects by juxtaposing the real and imaginary worlds.
Speech acts are simply things people do through language - for example, apologizing, complaining, instructing, agreeing and warning. The term ‘
act’ was coined by the linguistic philosopher
(1962) and developed by another philosopher
(1969). They maintained that, when using language, we not only make prepositional statements about objects, entities, states of affairs and so on, but we also fulfill functions such as requesting, denying, introducing, apologizing etc. Identifying the speech act being performed by a particular utterance can only be done if we know the context in which the utterance takes place. The functional intention of the speaker is known as the
of the utterance.
Using background knowledge: propositional level
With the insight that there is more to comprehending discourse than knowing the words on the page, have come attempts to provide theoretical models that can explain the ways in which our knowledge of the world guides our efforts to comprehend discourse. Much of this work has been carried out by researchers in the field of artificial intelligence. Their aim is to develop programs that will enable computers to comprehend and produce natural discourse.
Making sense of discourse
Using background knowledge: functional level
Background knowledge might help us interpret discourse on a functional level. When studying functions, the question is not 'what is the speaker/writer trying to tell us about events and things in the world?’ but ‘what is the speaker/writer trying to achieve through language?’
(1983) provides a lively piece of (fictional) interaction to demonstrate the points he wishes to make:
How we process discourse
Three models of how the comprehension process works
In the case of reading, the bottom-up model assumes that the reader first identifies each letter in a text as it is encountered. These letters are blended together and mentally ‘sounded out’ to enable the reader to identify the words that they make up.
Bottom-up processing seems a reasonable and logical explanation of what happens when we read; twenty-six written symbols have to represent over forty aural symbols, the correspondences between letters and sounds are both complex and relatively unpredictable. It seems more logical to teach beginning readers to exploit the systematic correspondences between written and spoken symbols than to teach them to recognize the words they encounter by memorizing each word’s unique shape, but later the processing of each letter as it is encountered in a text would slow reading down to the point where it would be difficult for meaning to be retained.
Top-down processing operates in the opposite direction from bottom-up processing: listeners/readers make sense of discourse by moving from the highest units of analysis to the lowest. The listener/reader makes use of his or her background knowledge of the subject at hand, knowledge of the overall structure of the text, knowledge and expectations of how language works, and motivation, interests and attitudes towards the text and the context it contains.
As the name indicates, this model suggests that, in comprehending discourse, we use information from more than one level simultaneously. In other words, comprehension is not a simple matter - neither of moving from lower to higher, nor from higher to lower elements - but is an interactive process.
In interactive models, deficiencies at one level can be compensated for by any other level, regardless of whether it is higher or lower in the hierarchy. For example, higher-level processes can make up for deficiencies at lower levels, and this allows for the possibility that readers with, say, poor reading skills can compensate for these by using other factors. These factors might include knowledge of the syntactic class of a given word or higher-level semantic knowledge.
Top-down strategies that good readers employ, and that can be taught to young readers, include the following:
• Using background knowledge to assist in comprehending a particular- text;
• Scanning the text for headings, sub-headings and non-text material such as pictures, graphs and diagrams to acquire a broad understanding before more detailed reading;
• Skimming the text and thinking about the content, and then writing down a number of questions that you would like the text to answer for you;
• Identifying the genre of the text (knowing that you are about to read a procedural, instructive, allegorical text etc. can facilitate reading comprehension);
• Discriminating between more and less important information (for example, discriminating between key information and supporting detail).
A: I have two tickets to the theatre tonight. – “Invitation”
B: My examination is tomorrow.- “Polite refusal”
Negotiating procedures depend crucially on the participants knowing what each utterance stands for functionally.
Conversation analysts attempt to describe and explain the ways in which conversations work. Their central question is: ‘How is it that conversational participants are able to produce intelligible utterances, and how are they able to interpret the utterances of others?’ This type of analysis is rather different from other forms of discourse analysis, the differences stemming in part from the fact that it was developed within a sociological rather than linguistic tradition, the school itself being known by the rather intimidating term ethnomethodology.
Ethnomethodologists insist that data should be derived from naturally occurring instances of everyday interaction. In a key collection of papers on conversation analysis, Atkinson and Heritage (1984) point out that virtually none of the data in their book: could conceivably be the product of recollection or intuition ... Such “invented” data can always be viewed as the implausible products of selective processes involving recollection, attention or imagination ... They argue that the success of an experiment will depend on the extent to which the researcher has been able to limit, control and manipulate the behaviour in question.
A characteristic feature of this type of research is the very elaborate analysis of relatively small samples of language.
Davidson (1984) investigates what happens in a conversation when an invitation, offer, request or proposal is rejected. Her database includes relatively short rejection sequences such as the following:
A: I was gonna say if you wanted to you could meet me at UBC and I could show you some of the other things on the computer, maybe even show you how to program Basic or something.
B: Well, I don‘t know if I’d want to get all that involved.
A: It’s really interesting.
(Adapted from Davidson 1984: 108)
Among other things, Davidson claims to have found that, following a rejection, speakers typically reformulate their offer, and that the subsequent version provides the interlocutor with an alternative which provides a face-saving way for the interlocutor to reject the offer.
Questions that conversation analysts have investigated include the following.
— How do topics get nominated, accepted, maintained and changed?
— How is speaker selection and change organized?
— How are conversational ambiguities resolved? .
— How are non-verbal and verbal aspects of conversation organized and integrated?
— What role does intonation play in conversation management?
— What recurring functional patterns are there in conversation, and how are these organized?
— How is socially sanctioned behaviour (for example, politeness versus rudeness, directness versus indirectness) mediated through language?
All but the most constrained interactions are the result of the joint efforts of the participants to make sense to each other. This is reflected in the amount of negotiation which is required in order for conversational participants to stay ‘on track’ - that is, for speakers to ensure that their messages are being received in the way they intended, and for the listeners to ensure that they are interpreting what they hear correctly.
A: How do I get to Kensington Road?
B: Well you go down Fullarton Road.. .
A: ... what, down Old Belair, and around. . .?
B: Yeah. And then you go straight...
A; ... past the hospital?
B: Yeah, keep going straight, past the racecourse to the roundabout.
You know the big roundabout?
B: And Kensington Road’s off to the right.
A: What, off the roundabout?
A: ... the achitecturai drawings will show us the elevations on the building.
Do you know what I mean by elevations?
B: elevations, levels?, no
A: it would be an elevation
B: Oh yeah front
A: from that elevation
B; Yeah, yeah, I understand
A: side elevation
In these extracts, there are numerous examples of the speakers and listeners doing conversational ‘work’ to ensure that there is mutual comprehension. In extract 1), the listener uses phrases such as the following to ensure that she has correctly understood the speaker: ... what, down Old Belair, and around ...?... past the hospital? What, off the roundabout?
At other points in the extracts the listeners signal comprehension by using phrases such as: Right! Yeah, yeah, I understand.
It is argued that when learners are put into a position where they have to negotiate meaning in order to make themselves comprehensible to their interlocutors, they will be pushed to the limits of their competence, and that this will ‘fuel’ the acquisition proccss. They have found that problem-solving and information gap tasks seem to stimulate the maximum amount of negotiation.
Researchers have shown that conversational dynamics and the performance of speech acts differ from language to language and culture to culture. Learning another language therefore involves much more than learning the pronunciation, vocabulary and grammar.
A study by
, for example, showed that when readers were exposed to written texts which described aspects of a culture foreign to them, there was a breakdown in comprehension.
Odlin (1989) provides a very useful survey of research into, crosslinguistic comparisons of discourse. He argues that investigating discourse is more challenging than other areas such as phonology or syntax, but that it is a critically important area for applied linguistic research: Cross-linguistic differences in discourse may affect comprehension as well as production.
Studies have been carried out into a number of different speech acts, including requests, apologies and greetings. All show considerable variation from one language to another.
Thank you for your attention!